CN106054195B - The turbulent flow spectrum width method of estimation of optimal processor during based on sky - Google Patents

Abstract

A method for estimating turbulent spectral width based on a space-time optimal processor. It includes modeling the turbulence echo of the airborne pulse Doppler weather radar under the phased array system, so as to obtain the weather radar echo data of the turbulence field; respectively constructing the generalized space steering vector and the generalized time for the turbulence field Steering vector, so as to obtain its space-time steering vector; combined with the space-time steering vector constructed in step 2), construct a space-time optimal processor to process radar echo data, suppress the interference caused by non-meteorological factors and ensure that the interference caused by turbulent targets The power of the radar echo remains unchanged, and the turbulent spectral width is estimated; the echo data of all range units within the radar working range are processed sequentially, and the speed spectral width estimation results of each range unit are estimated. The method of the invention can effectively suppress the spectrum spread interference caused by non-meteorological factors, and accurately estimate the turbulence spectrum width, and the simulation experiment has verified the validity of the method.

Description

Estimation method of turbulent spectral width based on space-time optimal processor

technical field

The invention belongs to the technical field of radar signal processing, in particular to a turbulence spectral width estimation method based on a space-time optimal processor.

technical background

Turbulence refers to continuous random fluctuations superimposed on the average wind. It is an atmospheric disturbance phenomenon often encountered during flight, and is usually caused by the rapid and irregular flow of the atmosphere. This kind of turbulence is easy to cause the aircraft to turbulence, and even make it greatly deviate from the scheduled route, so it is extremely detrimental to flight safety. The airborne weather radar can detect dangerous weather areas including turbulence, wind shear, thunderstorm, etc. in a certain sector in front of the aircraft route, and provide pilots with information such as the direction and intensity of harmful weather as a reference for early warning and avoiding dangerous areas .

For airborne weather radar, turbulence is a meteorological target with large particle velocity deviation. Velocity deviation can be understood as the fluctuation range or spectral width of velocity. The larger the spectral width, the greater the turbulence intensity. At present, turbulence detection is usually implemented by estimating the echo spectral width and comparing it with the detection threshold. It can be seen that the accuracy of the spectral width estimation result will directly affect the detection performance. Therefore, it is very necessary to improve the accuracy of turbulent spectral width estimation as much as possible for effective detection and early warning of turbulent dangerous meteorological regions.

In the process of turbulence detection, the movement of turbulent targets is not the only factor that causes spectrum expansion. Non-meteorological factors such as antenna azimuth and antenna beam width can also cause spectrum broadening, which affects the accuracy of real turbulence spectral width estimation results.

At present, the methods commonly used in spectral width estimation and turbulence detection mainly include Pulse-pair Processing (PPP) based on time-domain analysis and Fast Fourier Transformation (FFT) based on frequency-domain analysis. Although these methods are simple to calculate and have good performance under high SNR conditions, their spectral width estimation performance drops sharply when there is interference or low SNR, and these methods do not consider the spectrum caused by non-meteorological factors. wide spread, it is easy to cause an overestimation of the true spectral width of the turbulent flow.

Compared with the traditional single-antenna system, the phased-array pulse Doppler weather radar antenna is composed of multiple array elements, the phase of each array element is controllable, the beam pointing is flexible, and its echo signal contains the target spatial sampling information. By making full use of the space and time domain information of the signal, the spectrum spread caused by the radar scanning process can be adaptively suppressed, and the accurate detection of the target can be better realized.

Contents of the invention

In order to solve the above problems, the object of the present invention is to provide a method for estimating turbulent spectral width based on a space-time optimal processor.

In order to achieve the above object, the turbulent spectral width estimation method based on the space-time optimal processor provided by the present invention includes the following steps carried out in order:

1) Model the turbulence echo of the airborne pulse Doppler weather radar under the phased array system, so as to obtain the weather radar echo data of the turbulence field;

2) Based on the spatial and temporal distribution characteristics of the turbulent target, respectively construct the generalized space steering vector and the generalized time steering vector applicable to the turbulence field, so as to obtain its space-time steering vector;

3) Utilize the space-time adaptive processing principle, combined with the space-time steering vector constructed in step 2), construct a space-time optimal processor, process the radar echo data in step 1), suppress the interference caused by non-meteorological factors and ensure The power of the radar echo caused by the turbulent target is constant, and the turbulent spectral width is estimated;

4) Repeat steps 2) to 3), sequentially process the echo data of all range units within the radar working range, and estimate the speed spectrum width estimation results of each range unit.

In step 3), the space-time adaptive processing principle is used, combined with the space-time steering vector constructed in step 2), to construct a space-time optimal processor, to process the radar echo data in step 1), and to suppress non- The interference caused by meteorological factors keeps the power of the radar echo caused by the turbulent target unchanged at the same time, and the method of estimating the turbulent spectral width is: analyzing the factors that cause the expansion of the turbulent spectral width and its influence on the real turbulent spectral width estimation result, Construct a space-time optimal processor suitable for turbulent targets, filter radar echoes, and finally suppress the expansion of turbulent spectrum width caused by non-meteorological factors, while ensuring that the power of radar echoes caused by turbulent targets remains unchanged. The turbulent spectral width is estimated by using the non-coupling characteristics of Doppler frequency and turbulent spectral width.

The turbulent spectral width estimation method based on the space-time optimal processor provided by the present invention is aimed at the airborne weather radar of the phased array system, based on the distributed weather target characteristics of the turbulent flow, and uses the space-time adaptive processing principle to construct the optimal processor , to estimate the turbulent spectral width. The method of the invention can suppress the interference caused by non-meteorological factors, and more accurately estimate the turbulence spectrum width, and the simulation experiment verifies the effectiveness of the method.

Description of drawings

Figure 1 is a geometric observation diagram of turbulent flow.

Figure 2(a) and (b) are the space-time two-dimensional spectrograms of the radar turbulence signal, respectively, where Figure 2(a) is a top view and Figure 2(b) is a three-dimensional view.

Fig. 3 is the space-time domain response diagram of the space-time steering vector of the point target.

Fig. 4 is the space-time domain response diagram of the space-time steering vector of the distributed meteorological target.

Fig. 5 is the space-time two-dimensional spectrogram of the turbulent wind field echo of No. 75 distance unit.

Fig. 6 is a comparison chart of spectral width estimation results between the method of the present invention and the traditional pulse pair method.

Specific implementation method

The method for estimating the turbulent spectrum width based on the space-time optimal processor provided by the present invention will be described in detail below through specific examples.

The turbulent spectral width estimation method based on the space-time optimal processor provided by the present invention includes the following steps in order:

1) Model the turbulence echo of the airborne pulse Doppler weather radar under the phased array system, so as to obtain the weather radar echo data of the turbulence field;

Assume that the flight speed of the airborne pulse Doppler weather radar (hereinafter referred to as radar) is V _a , an N-element uniform linear array is placed along the vertical direction of the course, the pulse repetition frequency is f _r , the number of coherent processing pulses is K, and the emission pulse wavelength is lambda.

In the present invention, x _l represents the NK × 1-dimensional space-time snapshot data of the l (l=1, 2, ..., L) distance unit, and its expression is as follows:

x _l =s _l +n _l (1)

Among them, s _l and n _l represent the turbulent space-time snapshot and noise of the l-th distance unit respectively, assuming that the noise is additive white Gaussian noise.

For the turbulence field in the l-th distance unit, the radar sampling data can be written as an N×K matrix S _l . Among them, the element _{s l} (n,k) in the nth row and kth column of the matrix S l represents the nth (n=1,2,...N) array element of the radar, the kth (k=1,2,...K ) pulses to the sampling data of the lth distance unit, when there are Q meteorological scattering particles in the range of radar beam irradiation in the distance unit, the specific expression is as follows:

in and represent the spatial angular frequency and temporal angular frequency of the qth (q=1,2,…,Q) meteorological scattering particles respectively, θ _q , respectively represent the azimuth and elevation angles of the meteorological scattering particle relative to the radar, R _q is the slant distance between the qth meteorological scattering particle and the aircraft with the radar set, is the radar antenna receiving pattern, and v _q represents the radial velocity of the qth meteorological scattering particle relative to the radar.

Expand the above matrix S _l into a NK×1-dimensional column vector, which is the space-time snapshot s _l of the turbulence field. Then the echo signal in the radar full-range unit can be expressed as:

X＝[x ₁ x ₂ ... x _L ] ^T (3)

The Doppler velocity spectral width is the degree to which the Doppler velocity of different sizes in the radar beam irradiation range deviates from its average value. In fact, it is caused by the scattering particles having different radial velocities. The radial velocity v _q is dispersed in The vicinity of a certain center velocity is the main factor affecting the velocity spectrum width. However, in the process of radar scanning, when there is a certain broadening of the scanning angle, it will also cause spectrum expansion. If the resulting spectrum broadening is not considered, it will lead to overestimation of the true spectral width of turbulence.

As shown in Figure 1, the radar flies in a straight line along the X-axis at a constant flight speed V _a , and the radar antenna azimuth angle is α _a , then for a static scattering particle J within the beam irradiation range, its radial direction relative to the radar The velocity is v _q =V _a , and the Doppler frequency shift is:

Among them, α is the azimuth angle of the scattering particle J, and λ is the emission pulse wavelength. The spectrum spread caused by radar antenna beam width Δα and radar antenna azimuth angle α _a can be expressed as:

Using σ _a to represent the corresponding velocity spectrum width, then:

If the non-meteorological factors such as radar antenna beam width Δα, radar antenna azimuth angle α _a and the contribution of turbulence to the echo velocity spectrum width are approximately regarded as independent of each other, then the velocity spectrum width _σv of the radar echo in the turbulent field can be expressed as:

where σ _T ² represents the variance of the velocity spectrum of the turbulent flow. When processing the echo signal, if the spectrum broadening caused by non-meteorological factors such as radar antenna beamwidth Δα and radar antenna azimuth α _a is not considered during the radar scanning process, the estimated value of the velocity spectrum width σ _v of the radar echo As the turbulent spectral width, when σ _v >σ _T , it will cause an overestimation of the true spectral width of the turbulent flow. When performing turbulence detection, it is easy to cause false alarms. Therefore, the impact of the above-mentioned interference factors on the estimation results must be considered.

2) Based on the spatial and temporal distribution characteristics of the turbulent target, respectively construct the generalized space steering vector and the generalized time steering vector applicable to the turbulence field, so as to obtain its space-time steering vector;

a) The width of the radar main lobe is used as the prior information of the turbulence field within the radar irradiation range, and a generalized spatial steering vector suitable for distributed targets such as turbulence is established.

When the azimuth angle of the center of the radar main lobe direction is θ _i , the center elevation angle is , let the generalized spatial steering vector of the turbulence field within its irradiation range be Its expression is as follows

in, is the spatial guidance vector of the point target; For the deterministic angular signal density function, in the present invention, the turbulent target will be at the central azimuth angle θ _i and the central pitch angle The extensions on are expressed as:

in, σ _θ , Respectively represent the center azimuth angle θ _i and the center elevation angle The angular spread in the direction.

b) Based on the Gaussian distribution characteristics of meteorological echoes, a generalized time-steering vector suitable for distributed targets such as turbulence is constructed.

Turbulent radar echoes are superimposed by a large number of scattering particle echoes, each scattering particle has a random phase, and there is relative motion between the scattering particles, so the radar echo has spectrum expansion. According to the central limit theorem, the superposition of the scattering electric field of a large number of scattering particles can obtain a Gaussian statistical signal. Therefore, the power spectrum of meteorological echoes such as turbulence is generally modeled as a Gaussian spectrum, and a signal with a Gaussian distribution of the power spectrum can be obtained by introducing Gaussian attenuation into the time-domain Doppler signal. From this, the generalized time-oriented vector that can describe distributed meteorological objects such as turbulence field can be obtained:

s _t (f _d ,σ _f ) _K×1 ＝v _t (f _d )⊙g _t (σ _f ) (10)

Among them, f _d =2v/λ represents the Doppler frequency, v _t (f _d ) represents the time-steering vector of a point target whose radial velocity is v; σ _f represents the Doppler spectral width of the signal, g _t (σ _f ) represents the frequency spread function, which can be expressed as follows:

Further, the Kronecker product of the obtained generalized space steering vector and the generalized time steering vector of the turbulence field can be obtained to obtain its space-time steering vector:

3) Utilize the space-time adaptive processing principle, combined with the space-time steering vector constructed in step 2), construct a space-time optimal processor, process the radar echo data in step 1), suppress the interference caused by non-meteorological factors and ensure The power of the radar echo caused by the turbulent target is constant, and the turbulent spectral width is estimated;

Define the power factor as Z, and its expression is as follows:

Among them, w represents the weight vector of the optimal processor; w ^H R(f _d ,0)w and w ^H R(f _d ,σ _f )w respectively represent the Doppler spectral width σ _f of the signal at different values. The output power of the optimal processor, R(f _d ,σ _f ) represents the theoretical covariance matrix of the radar echo. When the Doppler spectral width σ _f =0 of the signal, R(f _d ,0) is only related to the Doppler frequency f _d , and the spectrum expansion at this time is caused by the joint action of non-meteorological factors during the radar scanning process, By minimizing the output power w ^HR (f _d ,0)w of this part of the echo, the interference caused by the joint effect of the radar antenna beam width Δα and the radar antenna azimuth angle α _a on the turbulence spectral width estimation result can be suppressed. R(f _d ,σ _f ) can be obtained by the following formula:

R(f _d ,σ _f )＝S(f _d ,σ _f )S ^H (f _d ,σ _f ) (14)

Find the weight vector w of the optimal processor. Under the condition that the output power of the turbulent target echo remains unchanged, the spectral width expansion caused by non-meteorological factors is minimized, which is equivalent to maximizing the power factor Z. At this time, the optimal An optimal processor can be described by the following mathematical optimization problem:

According to the generalized CAPON criterion, the weight vector of the optimal processor is obtained by solving:

w=p{R ^-1 (f _d ,0)R(f _d ,σ _f )} (16)

Among them, p{ } represents the eigenvector corresponding to the largest eigenvalue of the solution matrix. Using _xi to represent the received data of the turbulent field of the distance unit to be detected, the output signal of the optimal processor is:

y=w ^H x _i (17)

Convert the Doppler spectral width σ _f of the signal to the velocity spectral width σ _v =σ _f ·λ/2, then w=p{R- ¹ (f _d ,0)R(f _d ,σ _f )}. Since the Doppler average frequency is not coupled to the Doppler spectral width, the estimation of the average frequency can be performed independently of the spectral width, while the estimation of the spectral width must be performed in conjunction with the estimation of the average frequency. According to this idea, when solving the weight vector w of the optimal processor, any spectral width can be fixed first (C is a constant greater than zero, unit: m/s), estimate the Doppler frequency, and then estimate the Doppler spectrum width, so as to reduce the computational complexity.

Get the average Doppler frequency estimate for the range cell to be detected Afterwards, the Doppler spectral width can be estimated. When the power factor Z is maximized, it means that the optimal processor suppresses the interference factor and matches the turbulent signal best. The weight vector w of the optimal processor is obtained, and the output signal w ^H x _i corresponds to the maximum power point The spectral width is the estimated value of the Doppler spectral width of the turbulence signal in the distance unit to be detected, and the estimated result is:

4) Repeat steps 2) to 3), sequentially process the echo data of all range units within the radar working range, and estimate the speed spectrum width estimation results of each range unit.

The effect of the turbulent spectral width estimation method based on the space-time optimal processor provided by the present invention can be further illustrated by the following simulation results.

Simulation parameter setting: set the radar aircraft flight speed V _a =200m/s, flight altitude H=8000m, the turbulence field is distributed at 9-21km in front of the radar, the number of radar antennas is N=8, and the distance between array elements is d=λ /2 ideal uniform linear array, radar working wavelength λ=0.032m, number of coherent processing pulses K=16, pulse repetition frequency f _r =1500Hz, azimuth angle of 60°, elevation angle of 0°, beam width of 3° , the minimum resolvable distance is 150m, and the signal-to-noise ratio is 20dB.

Figure 2(a) and (b) are the space-time two-dimensional spectrograms (top view and three-dimensional view) of radar turbulence echoes, respectively. Turbulent flow is a distributed target, and the scattered particles in the turbulent flow field are dispersed in a large space range. It can be seen from Figure 2 that the echo signal has a certain expansion in the spatial distribution; at the same time, due to the The scattering particles move irregularly, the direction of velocity changes sharply, and the fluctuation range of velocity is large. The echo signal has a large expansion in the frequency distribution, which leads to the broadening of the Doppler spectrum.

Fig. 3 is the space-time domain response diagram of the space-time steering vector of a point target; Fig. 4 is the space-time domain response diagram of the space-time steering vector of distributed meteorological targets such as turbulence; Fig. 5 is the turbulence echo of No. 75 distance unit Space-time two-dimensional spectrum. It can be seen that the space-time steering vector for the turbulent distributed meteorological target proposed by the present invention can better fit the actual turbulence signal, and the resulting steering vector mismatch error is small.

Fig. 6 is a comparison chart of spectral width estimation results between the method of the present invention and the traditional pulse pair method. Under the same conditions, the traditional pulse pair method does not consider the spectral width expansion caused by the radar antenna beam width and radar antenna azimuth angle during the scanning process of the radar antenna, and the estimated result has a large deviation from the true value of the spectral width (the average deviation is about 0.55 m/s). And the method of the present invention has suppressed the spectral width expansion that non-meteorological factors cause before carrying out spectral width estimation, more accurate to the Doppler velocity spectral width estimation result of each range gate, and less (average deviation is about 0.05 with true value deviation) m/s), so it is superior to traditional methods.

Claims (2)

1. A turbulence spectral width estimation method based on space-time optimal processor, is characterized in that, described spectral width estimation method comprises the following steps carried out in order: 1) Model the turbulence echo of the airborne pulse Doppler weather radar under the phased array system, so as to obtain the weather radar echo data of the turbulence field; 2) Based on the spatial and temporal distribution characteristics of the turbulent target, respectively construct the generalized space steering vector and the generalized time steering vector applicable to the turbulence field, so as to obtain its space-time steering vector; 3) Utilize the space-time adaptive processing principle, combined with the space-time steering vector constructed in step 2), construct a space-time optimal processor, process the radar echo data in step 1), suppress the interference caused by non-meteorological factors and ensure The power of the radar echo caused by the turbulent target is constant, and the turbulent spectral width is estimated; 4) Repeat steps 2) to 3), sequentially process the echo data of all range units within the radar working range, and estimate the speed spectrum width estimation results of each range unit. 2. the turbulence spectral width estimation method based on space-time optimal processor according to claim 1, is characterized in that: in step 3), described utilization space-time adaptive processing principle, in conjunction with step 2) constructs space-time steering vector, construct a space-time optimal processor, process the radar echo data in step 1), suppress the interference caused by non-meteorological factors and keep the radar echo power caused by turbulent targets unchanged, and estimate The method of turbulent spectral width is: analyze the factors that cause the expansion of turbulent spectral width and its influence on the real turbulent spectral width estimation result, construct a space-time optimal processor suitable for turbulent targets, filter radar echoes, and finally The expansion of turbulent spectral width caused by non-meteorological factors is suppressed, while the power of radar echo caused by turbulent targets is kept constant, and the turbulent spectral width is estimated by using the uncoupling characteristics of Doppler frequency and turbulent spectral width.