CN107561508B - Coherent accumulation detection method for uniformly accelerated moving target - Google Patents

Abstract

The invention discloses a coherent accumulation detection method for a uniformly accelerated moving target, and belongs to the technical field of radars. The invention firstly utilizes PD radar to transmit linear frequency modulation signals, and then carries out pulse compression on received echo signals containing targets. And then carrying out coordinate shifting transformation on the echo signal after pulse compression so as to correct the distance walk of the target. Then, the fractional Fourier transform is used to eliminate the Doppler walk of the target and realize coherent accumulation of energy. The invention can effectively inhibit noise and improve the detection performance of the radar on weak targets because the invention simultaneously utilizes the amplitude and phase information in the target echo to carry out long-time coherent accumulation. In addition, the fractional Fourier transform can be realized through the fast Fourier transform, so that the operation amount can be greatly reduced, and the fast detection of the radar on the weak target is facilitated.

Description

Coherent accumulation detection method for uniformly accelerated moving target

Technical Field

The invention belongs to a moving target detection technology in the technical field of radars, and particularly relates to a coherent accumulation detection method for a uniformly accelerated moving target.

Background

In recent years, with the development and progress of scientific technology, how to quickly and effectively detect a moving target with a low signal-to-noise ratio becomes a challenging problem in the field of radar signal processing. Both long range targets and low radar cross-sectional area targets are often referred to as weak targets because of their generally low echo snr. To effectively detect a weak target, increasing the number of echoes accumulated and extending the accumulation time is an effective method. However, as the accumulation time is extended, range walking and doppler walking of the target inevitably occur. Therefore, the effect of compensating for the range walk and the doppler walk determines the effect of detecting a weak target.

At present, radar target echo long-time accumulation methods can be mainly divided into non-coherent accumulation and coherent accumulation. The non-coherent accumulation only considers the amplitude information of the echo data, and the realization mode is simpler. However, the cumulative gain of this type of method in a low signal-to-noise environment is greatly reduced because its phase information is not used. Coherent accumulation considers the amplitude and phase information of echo data at the same time, and particularly, echoes are superposed in phase, so that higher accumulation gain can be obtained.

Conventional coherent accumulation methods such as Moving Target Detection (MTD) techniques can accumulate and detect moving targets with dwell times within a range or doppler cell. When the target moves with distance, namely the target spans a plurality of distance units, the traditional accumulation method MTD fails. The distance walk can be corrected by methods such as Keystone transformation, AR-MTD algorithm, Radon Fourier transformation and the like, and energy accumulation and detection of the target moving at a constant speed are realized. However, for even acceleration motion, where the motion pattern is more complex, it can present the problem of the target crossing multiple range cells and multiple doppler resolution cells simultaneously, i.e. range walking and doppler walking occur simultaneously. Doppler walk can cause dispersion of the energy in the frequency domain during accumulation, making the above method ineffective.

To solve the range walk and the doppler walk, various methods are proposed. The two-step second-order Keystone transformation carries out distance walking correction and energy accumulation through repeated interpolation, however, the method has a large amount of interpolation loss, and multiple times of interpolation are complex and tedious in process. The generalized Radon Fourier transform can eliminate distance walking and Doppler walking, and coherent accumulation is carried out on target energy through three-dimensional joint search of distance, speed and acceleration. However, the computation of the generalized Radon fourier transform is very complex, affecting the real-time performance of radar signal processing. In general, the existing method is mainly realized by interpolation operation or multi-dimensional parameter search, the operation amount is larger, the realization mode is more complex, and the practicability is reduced.

Disclosure of Invention

The invention aims to: aiming at the existing problems, the distance walk and Doppler walk effects of the uniformly accelerated moving target can be corrected, and coherent accumulation of target energy and target detection are realized in a low signal-to-noise ratio environment.

The invention relates to a coherent accumulation detection method for a uniformly accelerated moving target, which comprises the following steps:

step 1: transmitting a chirp signal, denoted z, using PD (pulse Doppler) radar_t(τ,t_n) Denote the target echo signal received by the radar as z_r(τ,t_n). Where τ is the fast time, i.e., the time taken from transmission to reception of a single pulse (which may also be used to correspondingly represent the target distance); t is t_nIndicating slow time, i.e. time required for the nth pulse signal, i.e. t_nnT (N1.., N), T and N respectively denote a pulse repetition interval and a total number of pulses emitted.

Based on the initial distance s from the radar to the target₀The target radial velocity v and the radial acceleration a can be obtained, and the target and the radar are in t_nThe distance at the moment is:

step 2: pulse compression processing is carried out on the echo signal to obtain a pulse pressure echo signal, which is recorded as z_c(τ,t_n). To the pulse pressure echo signal z_c(τ,t_n) Middle fast time tau, slow time t_nPerforming discrete processing to obtain fast time m and slow time n after the discrete processing, and recording the echo signal in m-n domain after the discrete processing as z_c(m, n), wherein m ═ f_sτ、

f_sIs the sampling frequency.

And step 3: traversing search is carried out on angle variables in an angle search range based on set search step length delta sigma, variable substitution is carried out by utilizing a conversion angle sigma searched each time (namely coordinate shifting conversion is carried out on m and n, so the conversion angle sigma can also represent a shifting angle, one shifting angle corresponds to a target speed, namely the search of the shifting angle is equivalent to the search of the target speed), and fast time m 'and slow time n' after coordinate shifting conversion are obtained, so that an echo signal after coordinate shifting conversion is obtained as z_c(m ', n'; σ), wherein the coordinate shifting transformation formula is:

and for each time of the echo signal z after the coordinate shifting transformation substitution_c(m ', n'; sigma) performing fractional Fourier transform and amplitude accumulation, and taking the transformation angle sigma corresponding to the maximum accumulated amplitude as the estimation value of the angle variable and recording the value as

Namely, it is

Wherein FRFT (-) represents the fractional order fast Fourier transform; | · | represents a modulo operation;

indicates σ corresponding to the object in parentheses when the object takes the maximum value.

And 4, step 4: judging the current estimated value

Whether the error is within a preset error range (a value range for accuracy judgment is preset), if not, adjusting the search step length delta sigma and continuing to execute the step 3; if yes, the estimation of the angle variable is accurate, and the current angle variable estimation value is used

Substitution of the echo signal to complete the correction of the distance walk, i.e. to

As echo signals corrected for distance walk, wherein

For simplicity of description, the corrected echo signal is recorded as z_c(m ', n'). After correction of the range walk, all echoes are located

Within the distance cell, wherein,

representing the initial distance s to the radar to the target₀The corresponding distance unit. Distance unit

Is noted as z_c(n'). To z_c(n') eliminating Doppler walk by performing fractional Fourier transform, and performing coherent accumulation of energy.

And 5: comparing the peak value after the coherent accumulation with a preset threshold value, and when the accumulated peak value is higher than the threshold value, indicating that the target can be detected; otherwise the target cannot be detected.

The formula of the fractional Fourier transform in the step 4 is as follows:

wherein, F_p(. -) represents the fractional Fourier transform operation with p as a variable;

represents the rotation angle, and alpha is the transformation order; k_p(n', u) denotes a transformation kernel.

The invention firstly utilizes PD radar to transmit linear frequency modulation signals, and then carries out pulse compression on received echo signals containing targets. And then the echo signal after pulse compression is subjected to variable substitution to correct the distance walk of the target. Then, the fractional Fourier transform is used to eliminate the Doppler walk of the target and realize coherent accumulation of energy.

In summary, due to the adoption of the technical scheme, the invention has the beneficial effects that: meanwhile, the amplitude and phase information in the target echo are utilized to carry out long-time coherent accumulation, so that the noise can be effectively inhibited, and the detection capability of the radar on the uniformly accelerated moving target is improved. In addition, the fractional Fourier transform can be realized through the fast Fourier transform, so that the operation amount can be greatly reduced, and the fast detection of the radar on the weak target is facilitated.

Drawings

FIG. 1 is a block flow diagram of a method provided by the present invention;

FIG. 2 shows the result of target echo pulse compression;

FIG. 3 shows the distance walk correction results using the present invention;

fig. 4 shows coherent accumulation results using the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the following embodiments and accompanying drawings.

The method is mainly verified by a simulation experiment method, and the correctness of all the steps and conclusions is verified by scientific computing software Matlab R2014 a. Setting the number of the targets as 1, the initial distance of the targets relative to the radar as 300km, the radial speed of the targets as 2500m/s and the radial acceleration of the targets as 24m/s²The radar transmitting carrier frequency is 0.3GHz, the radar bandwidth is 5MHz, the radar pulse repetition frequency is 500Hz, the total pulse number in the coherent accumulation time is 512, and the signal-to-noise ratio is-10 dB. The following provides a specific implementation of the present invention with reference to fig. 1:

step 1: transmitting a chirp signal using a PD radar noted z_t(τ,t_n) Denote the target echo signal received by the radar as z_r(τ,t_n). Wherein τ represents a fast time; t is t_nRepresents a slow time;

set target and radar at t_nThe distance at the moment is:

wherein s is₀V and a are the initial radar-to-target distance, target radial velocity and radial acceleration, respectively.

Step 2: pulse compression processing is carried out on the echo signals, and the signals after pulse pressure is obtained are recorded as z_c(τ,t_n). Fig. 2 shows the severe range walk that occurs with target echo after pulse pressure. Then, the pulse pressure echo signal z is processed_c(τ,t_n) Variables τ and t in (1)_nPerforming discrete treatment, i.e.

Wherein f is_sIs the sampling frequency. The echo signal in m-n domain after discrete processing is recorded as z_c(m,n)。

And step 3: in the search range [ sigma ]_min,σ_max]In the method, a traversal search is carried out on an angle variable sigma by taking delta sigma as a search interval, wherein the sigma is_minIs the lower bound of the search range, σ_maxRepresenting its upper bound. Carrying out variable substitution by using the transformation angle searched each time (namely, assigning the search result of each time to the angle variable sigma), and recording the echo signal after variable substitution as z_c(m ', n'; σ), where m 'is the new fast time variable after the move and n' is the new slow time variable after the move. The echo signal z after each time of coordinate shifting substitution_c(m ', n'; sigma) performing fractional Fourier transform to accumulate and compare amplitudes, wherein the maximum accumulated amplitude is the estimation value of the angle variable and is recorded as

And 4, step 4: judging the current estimated value

Whether the error is within a preset error range or not, if not, adjusting the search step length delta sigma and continuing to execute the step 3; if so, it indicates that the angle variable is accurately estimated, and

as the echo signal corrected for the distance walk, the echo signal after correction is represented as z for the sake of simplifying the description_c(m ', n'). As shown in fig. 3, the target distance walk is corrected through coordinate shift substitution. After correction of the range walk, all echoes are located

Within the distance cell, wherein,

is the initial distance s from the radar to the target₀The corresponding distance unit. Distance unit

Is noted as z_c(n'). To z_c(n') eliminating Doppler motion by performing fractional Fourier transform, and accumulating and focusing energy, namely coherent accumulation of energy. The coherent integration of the target energies is shown in fig. 4.

And 5: comparing the peak value after the coherent accumulation with a preset threshold value, and when the accumulated peak value is higher than the threshold value, indicating that the target can be detected; otherwise the target cannot be detected.

Claims (1)

1. A coherent accumulation detection method for a uniformly accelerated moving target is characterized by comprising the following steps:

step 1: transmitting a chirp signal using a pulsed Doppler radar and receiving an echo signal, recording the transmitted signal as z_t(τ,t_n) Denote the received target echo signal as z_r(τ,t_n) Wherein τ represents a fast time; t is t_nRepresents a slow time;

step 2: for echo signal z_r(τ,t_n) Performing pulse compression to obtain pulse pressure echo signal, denoted as z_c(τ,t_n)；

To the pulse pressure echo signal z_c(τ,t_n) Middle fast time tau, slow time t_nPerforming discrete processing to obtain fast time m and slow time n after the discrete processing, and recording the echo signal in m-n domain after the discrete processing as z_c(m, n), wherein m ═ f_sτ、

f_sIs the sampling frequency, T is the pulse repetition interval;

and step 3: traversing search is carried out on angle variables in an angle search range based on the set search step length delta sigma, coordinate moving transformation is carried out on the fast time m and the slow time n by utilizing the transformation angle sigma obtained by each search, and the fast time m 'and the slow time n' after the coordinate moving transformation are obtained, so that the echo signal z after the coordinate moving transformation is obtained_c(m ', n'; σ), wherein the coordinate shifting transformation formula is:

and for each time of the echo signal z after the coordinate shifting transformation substitution_c(m ', n'; sigma) performing fractional Fourier transform and amplitude accumulation, and taking the transformation angle sigma corresponding to the maximum accumulated amplitude as the estimation value of the angle variable and recording the value as

And 4, step 4: judging the current estimated value

Whether the error is within a preset error range or not, if not, adjusting the search step length delta sigma and continuing to execute the step 3; if yes, estimating the current angle variable

Substitution of echo signalsz_c(m ', n'; σ) correction of distance walk and recording the echo signal after correction as z_c(m′,n′)；

After correction of the range walk, all echoes are located

Within the distance cell, wherein,

representing the initial distance s to the radar to the target₀Corresponding distance cell, the distance cell

Is noted as z_c(n') and for the echo signal z_c(n') performing fractional Fourier transform and then performing coherent accumulation of energy;

the fractional Fourier transform formula is as follows:

wherein, F_p(. -) represents the fractional Fourier transform operation with p as a variable;

represents the rotation angle, and alpha is the transformation order; k_p(n', u) represents a transformation kernel;

and 5: comparing the peak value after the coherent accumulation with a preset threshold value, and when the accumulated peak value is higher than the threshold value, indicating that the target can be detected; otherwise the target cannot be detected.

Images