CN105388465A - Sea clutter simulation method based on sea wave spectrum model - Google Patents

Abstract

The invention discloses a sea clutter simulation method based on a sea wave spectrum model and mainly solves the problem that it is difficult to reflect wave height details on a sea surface in the prior art. The realization scheme is characterized by 1) establishing the sea surface based on the sea wave spectrum model; 2) determining sea surface range under irradiation of a radar main wave beam; 3) selecting a rectangular frame in the sea surface range, and carrying out discretization on the sea wave in the frame to form a complex surface formed by splicing a plurality of equal-area triangle scattering units; 4) calculating attribute parameters and echo power of each triangle scattering unit; 5) establishing an echo model of the triangle scattering units according to the echo power; and 6) generating echo waves of the triangle scattering units of the sea surface in the rectangular frame according to the echo model, judging radar visibility of the triangle scattering units, calculating and superposing the echo waves of all visible triangle scattering units and generating sea clutter simulation data and analyzing the data. The method improves sea clutter simulation reliability and practicality, and can be used for target detection.

Description

Based on the sea clutter simulation method of wave spectrum model

Technical field

The invention belongs to signal processing technology field, specifically a kind of sea clutter simulation method, can be used for target detection.

Background technology

The emulation technology of radar sea clutter, experienced by from shallow to deep, goes from the simple to the complex, again slightly to thin process.Be sea clutter is regarded as the random series with certain probability density function PDF at first, adopt Monte-Carlo method to emulate.Classical sea clutter amplitude statistics model has that rayleigh distributed, logarithm just too distributes, Wei Buer distribution and K distribute.And the actual sea clutter run into not only has the statistical distribution characteristic in amplitude, meet again certain correlativity simultaneously, this just makes the data emulated will meet certain specific power spectrum characteristic or correlation properties and amplitude distribution characteristic simultaneously, more representative and the method for maturation has spherically invariant random processes SIRP and Zero memory nonlinearity transfrom method ZMNL, also has a kind of few stochastic differential equation method SDE.But these are a kind of empirical models from the sea clutter amplitude statistics model of statistical theory, fail to consider actual operating conditions and the environmental baseline of radar, directly can not apply to the clutter analysis of actual radar system.

So need to carry out for concrete radar operative scenario the relation that clutter simulation comes research model parameter and radar parameter and environmental parameter.Scattering unit division has been carried out by distance-Doppler in sea, ground by document " any attitude records PD radar three-dimensional land clutter algorithm research ", depending on ground sea clutter during working method has carried out modeling and simulation under adopting to Airborne Pulse Doppler Radar; Document " modeling and simulation of PD radar seeker sea clutter spectrum " is in conjunction with models such as the revised surface scattering coefficients of measured data, scattering unit division is carried out by equidistant ring in sea, modeling and simulation has been carried out to the radar seeker sea clutter be operated under pulse Doppler system.The former is in theory very accurate according to the scattering unit division methods of distance-Doppler, but adopts mathematical formulae to represent these scattering unit more complicated, and calculating clutter power need calculate dual-integration, implements more difficult; The latter is fairly simple according to the scattering unit division methods of equidistant ring, but scattering unit area increases along with the increase of distance, and extra large surface details is difficult to embody, and reduces the confidence level of sea clutter simulation.

Summary of the invention

The object of the invention is to the deficiency for above-mentioned prior art, propose a kind of sea clutter simulation method based on wave spectrum model, improve confidence level and the practicality of sea clutter simulation.

For achieving the above object, technical solution of the present invention comprises as follows:

(1) theoretical according to random seaway, with wave height equation η (x, y, t) any point (x on datum water level is described, y) at the extra large surface elevation of t, set up the Wave Model on surface, sea, the amplitude wherein in wave height equation is determined by ocean wave spectrum function;

(2) setting up antenna height is H, and beam angle is 2 ψ, and wave beam boresight grazing angle is the head-down radar of φ, any point (x on surface, sea under irradiating according to antenna main beam, y, z) with the position relationship of head-down radar, calculate this movement locus x ²cos ²ψ-2yHcos φ sin φ+y ²(cos ²ψ-cos ²φ)=H ²(sin ²φ-cos ²ψ), determine the scope on surface, sea: as φ < ψ, extra large surface range is hyperbolic curve and internal sea surface thereof; As φ=ψ, extra large surface range is para-curve and internal sea surface thereof; As φ > ψ, extra large surface range is that ellipse and internal sea thereof are surperficial;

(3) the extra large surface range determined, choose the rectangle frame that size is p × q, in this rectangle frame, horizontal sampling interval is got with this surface, rectangle frame inland sea arbitrfary point (x ₀, y ₀, z ₀) movement velocity v, azimuth angle alpha ₀, angle of pitch β ₀, calculate longitudinal sampling interval s δ at maintenance DOPPLER RESOLUTION and under meeting the unambiguous condition of Doppler, and be respectively with transverse and longitudinal interval get sampled point with on the surface, sea of s δ in rectangle frame, then formed a triangle scattering unit with every 3 sampled points, wherein c is the light velocity, f _sfor radar signal sample frequency, s is the initial position in y-axis projection of ripple door, and δ is segmentation interval, position angle, n is the umber of pulse in a coherent processing interval CPI, and p is the length that the start-stop of ripple door projects in y-axis, and q is longitudinal width of the extra large surface range determined;

(4) parameter of any one triangle scattering unit is calculated:

(4a) according to the position relationship of any one triangle scattering unit and radar, the distance R between this triangle scattering unit and radar is calculated, azimuth angle alpha and angle of pitch β, arrival bearing normal direction area A ₀, the wherein area A of triangle scattering unit ₀, by calculate;

(4b) according to radar equation, the echo power P of this triangle scattering unit is calculated _r;

(5) echo model of establishment step (4) intermediate cam shape scattering unit wherein s _tthe linear frequency modulation LFM signal that () is radar emission, ζ (t) is phase delay shake, represent imaginary number, f _dfor the Doppler frequency of this triangle scattering unit;

(6) according to ray tracing principle, the hiding relation of hiding relation between triangle scattering unit and triangle scattering unit self is judged:

If certain triangle scattering unit summit is α relative to the elevation angle of radar ₁, blocking unit summit relative to the elevation angle of radar is α ₂, work as α ₁>=α ₂time, this triangle scattering unit is blocked, and does not calculate its echo; Work as α ₁< α ₂time, this triangle scattering unit does not have occluded cell to block, then calculate the local incident angle α of this triangle scattering unit _l, then judge the hiding relation of this triangle scattering unit self:

Work as α _lduring > pi/2, this triangle scattering unit is blocked by self, does not calculate its echo;

Work as α _lduring≤pi/2, this triangle scattering unit is not blocked, and calculates and superposes the echo of all triangle scattering units be not blocked, producing the sea clutter emulated data of a coherent processing interval CPI echo, complete the simulation to sea clutter.

The present invention compared with prior art tool has the following advantages:

1. contacting of the sea clutter that extra large surface enhanced that the method that the present invention uses wave spectrum model to be combined with radar running parameter, environmental parameter emulates is simulated and sea situation environment, has practicality more;

2. because discrete extra large surface can regard the complex surface be spliced by many triangle scattering units as, and triangle scattering unit is projected as right-angle triangle bin at surface level, the present invention adopts triangle scattering unit to divide to sea level, mutually than ever based on the resolution element division methods of equidistant ring, Triangular patch belongs to equal area partition, the bin area divided can not increase with the increase of distance, and the surface, sea of therefore simulating more can embody unrestrained high details.This division methods is simpler than distance-Doppler resolution element division methods on mathematical notation simultaneously, does not need complicated dual-integration computing;

Experiment simulation shows: 1) the amplitude curve figure of the sea clutter emulated data of single pulse echo of the present invention and the amplitude curve figure amplitude trend of sea clutter measured data basically identical, illustrate that the sea clutter that the present invention produces has confidence level; 2) amplitude statistical distribution of the sea clutter emulated data of the same range unit of the present invention and conventional statistics distributed model K fitting of distribution goodness check the best, show that sea clutter simulation method of the present invention has practicality, a kind of approach of relation between Research statistics distributed model parameter and radar running parameter and environmental parameter is provided.

Accompanying drawing explanation

Fig. 1 is operational flowchart of the present invention;

Fig. 2 is the power density equi-energy divided method schematic diagram based on wave used in the present invention;

Fig. 3 be use in the present invention radar main beam irradiate under determine extra large surface range schematic diagram;

Fig. 4 is that sampled point schematic diagram is got on the surface, sea in rectangle frame used in the present invention;

Fig. 5 is the structure triangle scattering unit schematic diagram used in the present invention;

Fig. 6 is location diagram between the triangle scattering unit of the calculating triangle scattering unit design parameter used in the present invention and radar;

Fig. 7 be the present invention emulate not in the same time sea surface X-Y scheme;

Fig. 8 is the comparison diagram of sea clutter emulated data amplitude curve figure of the present invention and sea clutter measured data amplitude curve figure;

Fig. 9 is the sea clutter echo data analysis chart that the present invention emulates;

Embodiment:

The present invention combines the surface, Pierson-Moskowitz wave spectrum model emulation sea based on wind speed, according to all kinds of relative merits sea level being divided to scattering unit method, the triangle scattering unit division methods of homalographic is proposed, calculate and superpose the echo simulation sea clutter of all scattering units, radar running parameter and environmental parameter is also take into account, as factors such as antenna height, beam angle and wave beam boresight grazing angles during emulation.

Below in conjunction with the drawings and specific embodiments, the present invention is described in further detail.

With reference to Fig. 1, the sea clutter simulation method based on wave spectrum model of the present invention, comprises the steps:

Step 1, sets up the Wave Model on surface, sea.

Due to impacts such as wind speed, the change that height rises and falls can be there is in sea, therefore known extra large surface can not always be even curface, therefore can be theoretical according to random seaway, actual wave is regarded as the result of the sine-wave superimposed as different angular frequency, the different direction of propagation, different wave height and different initial phase.

1a) select wave frequencies spectral function S _Ω(ω) with directivity function D (Φ):

The suggestiveness form adopting International Towing Tank Conference ITTC to propose is:

Wave frequencies spectral function S _Ω(ω) select Pierson-Moskowitz wave spectrum model, Pierson-Moskowitz spectrum is the Wind Wave Spectra of the Developed state taking wind speed as parameter, and its expression formula is:

$S_{\mathrm{\&Omega;}}\left(\mathrm{\&omega;}\right)=0.0081g^2-{\mathrm{\&omega;}}^{-5}\exp [-0.74\left(\frac{\mathrm{\&omega;}{U}_w}{g}\right)],$

Wherein ω is the angular frequency of wave of the sea, and Φ is the direction of propagation of wave relative to x-axis, and g is acceleration of gravity, U _wfor the wind speed higher than surperficial 19.5m place, sea;

1b) according to wave frequencies spectral function S _Ω(ω) with directivity function D (Φ), ocean wave spectrum function S (ω, Φ) is calculated,

S(ω,Φ)＝S _Ω(ω)D(Φ)；

1c) calculate the simple sinusoidal wave amplitude in different angular frequency and direction: wherein Δ ω and ΔΦ are respectively the increment of ω and Φ;

1d) set up following three dimensional random wave wave height equation according to above-mentioned parameter:

$\mathrm{\&eta;}(x,y,t)=\underset{m=1}{\overset{M}{\&Sigma;}}\underset{l=1}{\overset{L}{\&Sigma;}}{A}_{m,l}cos(k_{m}x{\mathrm{cos\&Phi;}}_l+k_{m}y{\mathrm{sin\&Phi;}}_l-{\mathrm{\&omega;}}_{m}t+{\mathrm{\&epsiv;}}_{m,l})$

Wherein A _m,lrepresent the sinusoidal amplitude in different angular frequency and direction; k _mfor wave number, according to the dispersion relation of deep water, the angular frequency of wave of the sea and k _mx=k _mcos Φ _l, k _my=k _msin Φ _l; ε _m,lfor random phase angle, obey [0,2 π) in be uniformly distributed; L represents direction Segmentation Number, _lrepresent the maximum fractionation number in direction, the larger segmentation precision of its value is higher, l=1, and 2 ..., L; M represents angular frequency Segmentation Number, and M represents the maximum fractionation number of angular frequency, m=1, and 2 ..., M;

The segmentation of diagonal frequencies with reference to Fig. 2, because angular frequency can by 1a) in Pierson-Moskowitz frequency spectrum function S _Ω(ω) describe, therefore the equi-energy divided method diagonal frequencies based on wave power density can be adopted to split: first to selected angular frequency m segmentation, make the energy Δ E at each angular frequency interval equal, then determine the centre frequency of each sub-area;

Definition angular frequency accumulation spectrum is:

Have according to the concept of energy in part: $\mathrm{\&Delta;}E={\&Integral;}_{{\mathrm{\&omega;}}_m}^{{\mathrm{\&omega;}}_{m+1}}{S}_{\mathrm{\&Omega;}}\left(\mathrm{\&omega;}\right)d\mathrm{\&omega;}=E\left({\mathrm{\&omega;}}_{m+1}\right)-E\left({\mathrm{\&omega;}}_m\right)=b,$

Wherein b is constant, and Mb=E (∞), mb are the centre frequency of each sub-area, and angular frequency is split into ω _m=mb;

1e) according to sea surface certain point (x, y, z) projection on datum water level coordinate be (x, y), η (x, y, t) is then used to represent t point (x, y) the extra large surface elevation at place, therefore can show that the Wave Model on surface, sea, point (x, y, z) place is: z=η (x, y, t).

Step 2, determines that antenna main beam irradiates the scope on surface of plunging into the commercial sea.

There is backward scattered major part and depend on these 3 factors of antenna height, beam angle and wave beam boresight grazing angle in surface, radar illumination sea.

2a) setting up one, to set up antenna height be H, and beam angle is 2 ψ, and wave beam boresight grazing angle is the head-down radar of φ, and antenna main beam points to y-axis direction, and the intersection point of wave beam boresight and y-axis is for putting a C, then radar beam boresight direction is: wherein, R _afor radar site, i, j, k are respectively the vector of unit length in three coordinate axis;

Be 2b) any point in radar footprint on datum water level with reference to Fig. 3, P (x, y, 0), then radar points to the direction vector of this point and is: if radar beam is needle-like wave beam or conical beam, obtain vector equation:

Wherein represent vector with vector inner product, || represent calculating vector field homoemorphism value;

2c) according to above-mentioned vector equation, the track of calculation level P (x, y, 0):

Lycosφ+LHsinφ＝(x ²+y ²+H ²) ^1/2Lcosψ，

Can obtain after arrangement:

x ²cos ²ψ-2yHcosφsinφ+y ²(cos ²ψ-cos ²φ)＝H ²(sin ²φ-cos ²ψ)，

For given radar altitude H, given wave beam boresight grazing angle φ and given beam angle 2 ψ, as φ > ψ, the track of this point is hyperbolic curve; As φ=ψ, the track of this point is para-curve; As φ > ψ, the track of this point is oval, and the surface, sea that the track of this point and inside thereof comprise is the extra large surface range determined.

Step 3, in the extra large surface range determined, structure triangle scattering unit.

When radar signal processor is to above-mentioned surface, the sea sampling determined, receives only the echoed signal of Bo Mennei, so a rectangle frame will be chosen, sampled point is got on the surface, sea in rectangle frame region, structure scattering unit.

3a) choose rectangle frame:

With reference to Fig. 4, choose the rectangle frame that size is p × q, p is the length that the start-stop of ripple door projects in y-axis, and q is longitudinal width of the extra large surface range determined, sampled point is got on the surface, sea then in this rectangle frame scope; Wherein may exist some sea surface not in the extra large surface range determined but in rectangle frame, because antenna beam can be weighted weakening to this part echo, this part extra large surface can not affect emulation.

3b) the lateral separation of calculating sampling point:

If the coordinate of radar is (x _r, y _r, z _r), the coordinate of extra large surperficial any point is (x ₀, y ₀, z ₀), antenna main beam points to y-axis direction, and radar signal processor is equidistantly sampled along the y-axis direction on datum water level, and the lateral separation of calculating sampling point is:

$\mathrm{\&Delta;}R=\frac{c}{2f_s}{\mathrm{sec\&phi;}}_0=\frac{c}{2f_s}\sqrt{1+{\tan }^2{\mathrm{\&phi;}}_0}=\frac{c}{2f_s}\sqrt{1+{(H/y_0)}^2},$

Wherein c is the light velocity, f _sfor sample frequency, φ ₀for the grazing angle at this some place, φ _s< φ ₀< φ _e, φ _s, φ _ebe respectively the grazing angle of ripple door section start, Bo Men end,

Point (x ₀, y ₀, z ₀) from the distance R of radar ₀for:

$R_0=\sqrt{{(x_R-x_0)}^2+{(y_R-y_0)}^2+{(z_R-z_0)}^2},$

Grazing angle is:

Along with y ₀increase, this point is from distance by radar R ₀increase, this some place grazing angle φ ₀reduce, Δ R reduces, and works as y ₀during > > H, the lateral separation of sampled point can abbreviation be:

3c) the longitudinal separation of calculating sampling point:

3c1) calculation level (x ₀, y ₀, z ₀) relative to the radial velocity of radar: v _r=vcos α ₀cos β ₀,

Wherein v is the movement velocity of scattering unit, α ₀for position angle,

β ₀for the angle of pitch, ${\mathrm{\&beta;}}_0=arccos\sqrt{\frac{{(x_R-x_0)}^2+{(y_R-y_0)}^2}{R_0}};$

3c2) calculation level (x ₀, y ₀, z ₀) Doppler frequency: wherein f _cfor signal carrier frequency;

3c3) calculation level (x ₀, y ₀, z ₀) azimuth resolution:

By 3c2) in Doppler frequency f _dto α ₀differential, namely obtaining azimuth resolution is: $d\left({\mathrm{\&alpha;}}_0\right)=\left|\frac{d\left(f_D\right)\&CenterDot;c}{2v{\mathrm{sin\&alpha;}}_0{\mathrm{cos\&beta;}}_0{f}_c}\right|,$ Wherein d () is infinitesimal;

3c4) computer azimuth angle segmentation interval:

The DOPPLER RESOLUTION of radar signal is wherein N is the umber of pulse in a coherent processing interval CPI, T _rfor the pulse repetition time, in order to keep DOPPLER RESOLUTION, d (f should be had _d)≤Δ f _d, get d (f _d)=Δ f _d, ${\mathrm{\&alpha;}}_0=\frac{\mathrm{\&pi;}}{2};$

V gets and makes the maximal value of Doppler not in fuzzy ranges, if ambiguous Doppler frequency range is not then the maximum occurrences of v meets: by above-mentioned d (f _d), α ₀, v parameter is updated to 3c3) in azimuth resolution d (α ₀) in, calculate position angle segmentation and be spaced apart:

For reaching not ambiguous Doppler frequency if the span of N is: N>=2; For making the division of δ meticulous as far as possible, can carry out getting for r time a little, r=1,2 ..., obtain { β ₁, β ₂..., β _r, get β ₀=min{ β ₁, β ₂..., β _r;

3c5) according to 3c4) in position angle segmentation interval δ and the coordinate s that projects in y-axis of the initial distance of ripple door, the longitudinal separation calculating extra large surperficial sampled point is s δ;

Sampled point is got on surface, sea 3d) in rectangle frame:

With reference to Fig. 5, with lateral separation sampled point is got with on the surface, sea of longitudinal separation s δ in rectangle frame, the black round dot that be projected as Fig. 5 (a) shown in of sampled point on datum water level, a triangle scattering unit is formed again with every 3 sampled points, surface, sea in rectangle frame is divided into the complex surface be spliced by many triangle scattering units, the right-angle triangle that be projected as Fig. 5 (b) shown in of scattering unit on datum water level.

Step 4, calculates parameter and the echo power of any one triangle scattering unit.

With reference to any one triangle scattering unit of Fig. 6 and radar site relation, this triangle scattering unit is at certain right-angle triangle be projected as in Fig. 5 (b) of surface level.The parameter of this triangle scattering unit comprises: the distance R between triangle scattering unit and radar, azimuth angle alpha and angle of pitch β, arrival bearing normal direction area A ₀.

4a) calculate the parameter of triangle scattering unit:

Three of any one triangle scattering unit apex coordinates are expressed as (x ₁, y ₁, z ₁), (x ₂, y ₂, z ₂), (x ₃, y ₃, z ₃), its center position coordinates is:

$(x_C,y_C,z_C)=\frac{1}{3}(x_1+x_2+x_3,y_1+y_2+y_3,z_1+z_2+z_3);$

4a1) calculate the distance between this triangle scattering unit and radar:

$R=\sqrt{{(x_R-x_C)}^2+{(y_R-y_C)}^2+{(z_R-z_C)}^2};$

4a2) calculate the direction (α, β) of this triangle scattering unit under radar antenna coordinate system:

Azimuth angle alpha: $\mathrm{\&alpha;}=arccos\frac{x_R-x_C}{\sqrt{{(x_R-x_C)}^2+{(y_R-y_C)}^2}},$

Angle of pitch β: $\mathrm{\&beta;}=arccos\frac{\sqrt{{(x_R-x_C)}^2+{(y_R-y_C)}^2}}{R};$

4a3) calculate the arrival bearing of this triangle scattering unit:

4a4) calculate the normal direction of this triangle scattering unit: $\stackrel{\&RightArrow;}{n_S}=\left|\begin{array}{ccc}i& j& k\\ x_2-x_1& y_2-y_1& z_2-z_1\\ x_3-x_1& y_3-y_1& z_3-z_1\end{array}\right|;$

4a5) calculate the area of this triangle scattering unit:

4b) calculate the echo power of this triangle scattering unit:

4b1) calculate this triangle scattering unit local grazing angle: its effective span is

4b2) according to the computing formula that Guinard and Dalcy provides, calculate the surface scattering coefficient Γ under horizontal polarization and vertical polarization _hHwith Γ _vV:

${\mathrm{\&Gamma;}}_{HH}=\frac{3}{2}\mathrm{\&pi;}\&CenterDot;{10}^{-3}{\left|\frac{\mathrm{\&epsiv;}-1}{{({\mathrm{sin\&alpha;}}_L+{\left(\right|\mathrm{\&epsiv;}-{\cos }^2{\mathrm{\&alpha;}}_L\left|\right)}^{1/2})}^2}\right|}^2{\tan }^4{\mathrm{\&alpha;}}_L,$

${\mathrm{\&Gamma;}}_{VV}=\frac{3}{2}\mathrm{\&pi;}\&CenterDot;{10}^{-3}{\left|\frac{(\mathrm{\&epsiv;}-1)\&lsqb;\mathrm{\&epsiv;}({\cos }^2{\mathrm{\&alpha;}}_L+1)-{\cos }^2{\mathrm{\&alpha;}}_L\&rsqb;}{{({\mathrm{\&epsiv;sin\&alpha;}}_L+{\left(\right|\mathrm{\&epsiv;}-{\cos }^2{\mathrm{\&alpha;}}_L\left|\right)}^{1/2})}^2}\right|}^2{\tan }^4{\mathrm{\&alpha;}}_L,$

Wherein ε is complex permittivity: ε=ε '-j ε ", wherein the value that for imaginary number, the single order of ε ' expression ε is led, ε " representing that the second order of ε is led, ε ' and ε " obtains by tabling look-up,

Table 1 gives the ε ' of some seawater complex permittivities when different frequency and temperature and ε " value:

The complex permittivity of table 1 seawater

The Radar Cross Section of corresponding this triangle scattering unit of formulae discovery 4b3) is selected according to the polarization mode of antenna:

During antenna horizontal polarization, the Radar Cross Section of this triangle scattering unit: σ=Γ _hHΑ ₀,

During antennas orthogonal polarization, the Radar Cross Section of this triangle scattering unit: σ=Γ _vVΑ ₀;

4b4) calculate this triangle scattering unit echo power according to above-mentioned parameter:

Wherein P _tfor radar transmission power, G (α _f, β _f; α, β) be wave beam boresight sensing (α _f, β _f) time (α, β) direction on antenna gain, λ is signal wavelength, L _sfor radar transmit-receive composite loss.

Step 5, sets up the echo model of triangle scattering unit.

5a) calculate radar transmitting wave:

If radar transmitting wave is chirp pulse signal:

s _t(t)＝rect(t/T)exp(j2πf _ct)exp(jπΒt ²)，

Wherein T is pulse width, and B is signal frequency modulation rate, $rect(t/T)=\left\{\begin{array}{c}1,\left|t\right|\&le;T/2\\ 0,\left|t\right|>T/2\end{array}\right.;$

5b) calculate the point scattering echoed signal of triangle scattering unit:

Wherein f _dfor the Doppler frequency of triangle scattering unit, can be estimated two interpulse changes of the distances relative to radar by triangle scattering unit, Δ t is the mistiming of two pulses;

5c) set up the echo model of triangle scattering unit:

Triangle scattering unit belongs to surface scattering unit, it can not be smooth triangle scattering surface, therefore the contribution of texture component to radar return on sea will be considered, to 3b) in point scattering echo model phase delay add and the effect of jitter simulation texture component obtain the echo model of triangle scattering unit:

$s_r\left(t\right)=\sqrt{P_r}\&CenterDot;s_t(t-\frac{2R}{c}+\mathrm{\&xi;}\left(t\right))\exp \left(j2{\mathrm{\&pi;f}}_{d}t\right),$

Wherein ζ (t) is phase delay shake, be average is zero, variance is σ ₀white Gaussian noise, represent the severe degree of shake, it is relevant with sea situation.

Step 6, judges the hiding relation of hiding relation between triangle scattering unit and triangle scattering unit self.

In reality, can exist and block between the triangle scattering unit of sea, triangle scattering unit itself also can exist and self blocks, and adopts ray trace and following principle and criterion to judge:

If certain triangle scattering unit summit is α relative to the elevation angle of radar ₁, blocking unit summit relative to the elevation angle of radar is α ₂, work as α ₁>=α ₂time, this triangle scattering unit is blocked, and does not calculate its echo; Work as α ₁< α ₂time, this triangle scattering unit does not have occluded cell to block, then calculate the local incident angle α of this triangle scattering unit _l, then judge the hiding relation of this triangle scattering unit self:

Work as α _lduring > pi/2, this triangle scattering unit is blocked by self, does not calculate its echo;

Work as α _lduring≤pi/2, triangle scattering unit is not blocked, and calculates and superposes the echo of all triangle scattering units be not blocked, producing the sea clutter emulated data of a coherent processing interval CPI echo, complete the simulation to sea clutter.

Step 7, analyzes the emulated data of sea clutter.

7a) the sea clutter emulated data of the coherent processing interval CPI echo produced in step 6 being stored in size is in the matrix Q of u × v, and wherein u is the umber of pulse that a CPI comprises, and v is the number of range unit;

7b) change antenna height parameter, produce two groups of sea clutter emulated datas, be respectively Q ₁, Q ₂, often organize and all get u ₁individual pulse, as the sea clutter emulated data of single pulse echo, is respectively wherein u ₁indicating impulse number; Draw respectively with emulation amplitude-range unit curve map, compare the amplitude of two curves;

7c) according to the sea clutter measured data Q of the single pulse echo of chirp-pulse compression radar collection _redraw actual measurement amplitude-range unit curve map, will q _rethe curve map of three is made comparisons, and analyzes emulation amplitude-range unit curve map and distributes with the amplitude trend of actual measurement amplitude-range unit curve map;

7d) produce many group sea clutter emulated datas, optional one group of Q ₃, draw two-dimentional reflectogram, select v wherein ₁the sea clutter emulated data of individual range unit draw its amplitude-pulse curve figure and amplitude statistical distribution curve map, its amplitude statistical distribution curve and conventional statistics distributed model curve are carried out the least mean-square error MSD test of fitness of fot, wherein v ₁represent range unit number.

Beneficial effect of the present invention is further illustrated below in conjunction with concrete the simulation results.

1. simulated conditions

As listed in table 2:

Table 2 simulation parameter

Emulation 1: under table 2 condition, emulate surface, the sea wave of slow time, result is as Fig. 7.Fig. 7 shows the surperficial X-Y scheme in emulation sea of 4 not 100m × 100m in the same time, shows emulate the dynamic change of the surperficial wave that goes to sea by surface, the sea Wave Model that the present invention sets up.

Emulation 2: add the condition that antenna height is respectively H=30m and H=500m under table 2 condition, extra large surface emulates sea clutter, the sea clutter emulated data of com-parison and analysis different antennae height and the emulated data of sea clutter and measured data, result is as Fig. 8.Wherein:

Emulation amplitude-range unit curve map that Fig. 8 (a) is the sea clutter of single pulse echo when antenna height is 30m;

Sea clutter emulation amplitude-range unit curve map that Fig. 8 (b) is single pulse echo when antenna height is 500m;

Sea clutter actual measurement amplitude-range unit curve map that Fig. 8 (c) is single pulse echo, wherein the measured data of sea clutter is collected by linear frequency modulation LFM pulse compression radar, this radar is positioned at certain irradiation Pacific Ocean, mountain top, antenna height is 500m, grazing angle is about 1 °, bandwidth 5MHz, pulsewidth 42 μ s;

Comparison diagram 8 (a) is known with Fig. 8 (b), Fig. 8 (b) echo amplitude comparatively Fig. 8 (a) echo amplitude is slightly large, shows the conclusion that sea clutter that the present invention produces meets sea clutter backscattering coefficient and increases with the increase of wave beam boresight grazing angle;

Fig. 8 (a) presents the low trend in middle high both sides with the emulation amplitude-range unit curve map in 8 (b) viewed from entirety, consistent with the actual measurement amplitude-range unit curve map trend in 8 (c), this is because Bo Menkai get is less, the scattering unit echo being in Bo Men both sides does not comprise afferent echo door completely.

Fig. 8 shows that the sea clutter that the present invention produces has confidence level.

Emulation 3: adding fixed antenna height under table 2 condition is again 500m, wave beam boresight grazing angle is the condition of 1 °, and emulation sea clutter, carries out the least mean-square error MSD test of fitness of fot to the amplitude statistical distribution of sea clutter and conventional statistics distributed model, result as Fig. 9, wherein:

Fig. 9 (a) is the sea clutter artificial echo X-Y scheme at a coherent processing interval;

Fig. 9 (b) is the sea clutter echo simulation amplitude-pulse curve figure of the 500th range unit;

The amplitude statistical distribution of Fig. 9 (c) for emulation amplitude-pulse curve figure and the fitting result of conventional statistics distributed model;

Fig. 9 (d) is the doppler spectral of the 500th range unit echo simulation data;

As shown in table 3 to emulating the result adopting the least mean-square error MSD test of fitness of fot to obtain in Fig. 9 (c), best by the visible K fitting of distribution result of table 3, rayleigh distributed fitting result takes second place;

Table 3 statistical distribution fitting result

The doppler spectral of the 500th range unit emulated data in Fig. 9 (d) has certain frequency displacement and spectrum width, because triangle scattering unit is different relative to the position angle of radar on emulation sea, and changes different from the Distance geometry distance between radar; The speed of triangle scattering unit that doppler spectral can be regarded as by antenna main beam irradiates sea maps, and Fig. 9 (c) is consistent with conventional statistics modeling results with Fig. 9 (d).

Fig. 9 result shows: sea clutter simulation method of the present invention has practicality, can be used for Research statistics model parameter, relation between radar running parameter and environmental parameter.

Claims (3)

1., based on the sea clutter simulation method of wave spectrum model, comprising:

(1) theoretical according to random seaway, with wave height equation η (x, y, t) any point (x on datum water level is described, y) at the extra large surface elevation of t, set up the Wave Model on surface, sea, the amplitude wherein in wave height equation is determined by ocean wave spectrum function;

(2) setting up antenna height is H, and beam angle is 2 ψ, and wave beam boresight grazing angle is the head-down radar of φ, any point (x on surface, sea under irradiating according to antenna main beam, y, z) with the position relationship of head-down radar, calculate this movement locus x ²cos ²ψ-2yHcos φ sin φ+y ²(cos ²ψ-cos ²φ)=H ²(sin ²φ-cos ²ψ), determine the scope on surface, sea: as φ < ψ, extra large surface range is hyperbolic curve and internal sea surface thereof; As φ=ψ, extra large surface range is para-curve and internal sea surface thereof; As φ > ψ, extra large surface range is that ellipse and internal sea thereof are surperficial;

(3) the extra large surface range determined, choose the rectangle frame that size is p × q, in this rectangle frame, horizontal sampling interval is got with this surface, rectangle frame inland sea arbitrfary point (x ₀, y ₀, z ₀) movement velocity v, azimuth angle alpha ₀, angle of pitch β ₀, calculate longitudinal sampling interval s δ at maintenance DOPPLER RESOLUTION and under meeting the unambiguous condition of Doppler, and be respectively with transverse and longitudinal interval get sampled point with on the surface, sea of s δ in rectangle frame, then formed a triangle scattering unit with every 3 sampled points, wherein c is the light velocity, f _sfor radar signal sample frequency, s is the initial position in y-axis projection of ripple door, and δ is segmentation interval, position angle, n is the umber of pulse in a coherent processing interval CPI, and p is the length that the start-stop of ripple door projects in y-axis, and q is longitudinal width of the extra large surface range determined;

(4) parameter of any one triangle scattering unit is calculated:

(4a) according to the position relationship of any one triangle scattering unit and radar, the distance R between this triangle scattering unit and radar is calculated, azimuth angle alpha and angle of pitch β, arrival bearing normal direction area A ₀, the wherein area A of triangle scattering unit ₀, by calculate;

(4b) according to radar equation, the echo power P of this triangle scattering unit is calculated _r;

(5) echo model of establishment step (4) intermediate cam shape scattering unit wherein s _tthe linear frequency modulation LFM signal that () is radar emission, ζ (t) is phase delay shake, represent imaginary number, f _dfor the Doppler frequency of this triangle scattering unit;

(6) according to ray tracing principle, the hiding relation of hiding relation between triangle scattering unit and triangle scattering unit self is judged:

If certain triangle scattering unit summit is α relative to the elevation angle of radar ₁, blocking unit summit relative to the elevation angle of radar is α ₂, work as α ₁>=α ₂time, this triangle scattering unit is blocked, and does not calculate its echo; Work as α ₁< α ₂time, this triangle scattering unit does not have occluded cell to block, then calculate the local incident angle α of this triangle scattering unit _l, then judge the hiding relation of this triangle scattering unit self:

Work as α _lduring > pi/2, this triangle scattering unit is blocked by self, does not calculate its echo;

Work as α _lduring≤pi/2, this triangle scattering unit is not blocked, and calculates and superposes the echo of all triangle scattering units be not blocked, producing the sea clutter emulated data of a coherent processing interval CPI echo, complete the simulation to sea clutter.

2. method according to claim 1, wherein set up the Wave Model on surface, sea in step (1), carry out as follows:

(1a) theoretical according to random seaway, obtain the mathematical description of three dimensional random wave wave height equation $\mathrm{\&eta;}(x,y,t)=\underset{m=1}{\overset{M}{\&Sigma;}}\underset{l=1}{\overset{L}{\&Sigma;}}{A}_{m,l}cos(k_m{\mathrm{xcos\&Phi;}}_l+k_m{\mathrm{ysin\&Phi;}}_l-{\mathrm{\&omega;}}_{m}t+{\mathrm{\&epsiv;}}_{m,l}),$ Wherein m represents angular frequency Segmentation Number, and M represents the maximum fractionation number of angular frequency, and l is direction Segmentation Number, and L represents the maximum fractionation number in direction; k _mfor the wave number of wave of the sea, ω _mfor the angular frequency of wave of the sea, ε _m,lfor random phase angle, Φ _lfor wave is relative to the direction of propagation of x-axis; A _m,lfor the simple sinusoidal wave amplitude in different angular frequency and direction, s (ω _m, Φ _l) be the Wave energy distribution in different frequency and direction, S (ω, Φ)=S _Ω(ω) D (Φ), S _Ω(ω) be wave frequencies spectral function, select Pierson-Moskowitz wave spectrum model here: g is acceleration of gravity, U _wfor the wind speed higher than surperficial 19.5m place, sea; D (Φ) is directivity function, adopts the suggestiveness form that International Towing Tank Conference ITTC proposes:

(1b) by upper for surface, sea any point (x, y, z) datum water level (x, y) point is projected on, with wave height equation η (x, y, t) describe the extra large surface elevation z of point (x, y) in t, the Wave Model setting up surface, sea is: z=η (x, y, t).

3. method according to claim 1, wherein step (4) calculates the echo power P of this triangle scattering unit _r, be calculated as follows:

$P_r=\sqrt{\frac{P_t{\mathrm{\&lambda;}}^2\mathrm{\&sigma;}}{{\left(4\mathrm{\&pi;}\right)}^3{L}_s}}\&CenterDot;\frac{G({\mathrm{\&alpha;}}_F,{\mathrm{\&beta;}}_F;\mathrm{\&alpha;},\mathrm{\&beta;})}{R^2},$

Wherein P _tfor radar transmission power, λ is signal wavelength, L _sfor radar transmit-receive composite loss, G (α _f, β _f; α, β) be wave beam boresight sensing (α _f, β _f) time (α, β) direction on antenna gain, σ is the Radar Cross Section of triangle scattering unit, and it is by σ=Γ _hHa ₀or σ=Γ _vVa ₀calculate, Γ _hHfor the surface scattering coefficient under horizontal polarization, ${\mathrm{\&Gamma;}}_{HH}=\frac{3}{2}\mathrm{\&pi;}\&CenterDot;{10}^{-3}|\frac{\mathrm{\&epsiv;}-1}{{({\mathrm{sin\&alpha;}}_L+{\left(\right|\mathrm{\&epsiv;}-{\cos }^2{\mathrm{\&alpha;}}_L\left|\right)}^{1/2})}^2}{|}^2{\tan }^4{\mathrm{\&alpha;}}_L,$ Γ _vVfor the surface scattering coefficient under vertical polarization, ${\mathrm{\&Gamma;}}_{VV}=\frac{3}{2}\mathrm{\&pi;}\&CenterDot;{10}^{-3}|\frac{(\mathrm{\&epsiv;}-1)\&lsqb;\mathrm{\&epsiv;}({\cos }^2{\mathrm{\&alpha;}}_L+1)-{\cos }^2{\mathrm{\&alpha;}}_L\&rsqb;}{{({\mathrm{\&epsiv;sin\&alpha;}}_L+{\left(\right|\mathrm{\&epsiv;}-{\cos }^2{\mathrm{\&alpha;}}_L\left|\right)}^{1/2})}^2}{|}^2{\tan }^4{\mathrm{\&alpha;}}_L,$ ε is complex permittivity, α _lfor the local grazing angle of triangle scattering unit, it is by this triangle scattering unit arrival bearing with normal direction determine: ${\mathrm{\&alpha;}}_L=\frac{\mathrm{\&pi;}}{2}-arccos\left(\frac{\stackrel{\&RightArrow;}{n_W}\&CenterDot;\stackrel{\&RightArrow;}{n_S}}{\left|\stackrel{\&RightArrow;}{n_W}\right|\&CenterDot;\left|\stackrel{\&RightArrow;}{n_S}\right|}\right).$