Efficient maneuvering weak target long-time coherent accumulation detection method

Abstract

The invention discloses an efficient maneuvering weak target long-time coherent accumulation detection method, and belongs to the field of radar target detection. Comprising the following steps: performing three-scale decomposition on maneuvering target radial motion parameter variables; eliminating a Doppler frequency ambiguity effect in a target ambiguity scale motion parameter variable search space; correcting first-order range migration caused by the baseband speed of the maneuvering target through Keystone transformation; estimating maneuvering target coarse-scale baseband motion parameters in the target coarse-scale baseband motion parameter variable search space through GIFT (Geometric Invariant Feature Transform); estimating fine-scale baseband motion parameters of the maneuvering target in a target fine-scale baseband motion parameter variable search space to obtain a temporary distance-Doppler spectrum; according to the maneuvering target motion parameter variable three-scale decomposition model, complete motion parameters of the maneuvering target are obtained through estimation; and a maneuvering target distance-Doppler spectrum is obtained based on maneuvering target motion parameter estimation. The method has the characteristics of low calculation complexity, effective Doppler ambiguity compensation, high detection performance and the like.

Description

Technical Field

[0001] The invention relates to the field of radar target detection, and in particular to an efficient long-time coherent accumulation detection method for maneuverable weak targets. Background Art

[0002] With the continuous development of maneuvering targets and the increasing maturity of stealth technology, efficient detection of moving radar targets has always been a difficult problem in the field of radar signal processing. For the detection of maneuvering weak targets, long-term coherent accumulation technology can effectively improve the detection performance by accumulating target energy through long-term observation. However, the relative motion between the target and the radar can cause its echo envelope peak to appear in different distance units, which is called the range migration effect; the relative acceleration between the target and the radar will cause the echo phase to change, causing the Doppler frequency to cross multiple Doppler units, which is called the Doppler migration effect. Both range migration and Doppler migration will cause the target energy to defocus and affect the detection performance. Therefore, it is necessary to estimate and compensate for motion parameters such as distance, velocity and acceleration.

[0003] The existing long-term coherent accumulation detection technology solutions for high-order RM and DFM are divided into the following three categories: 1) Keystone transform (KT) type methods: KT corrects the linear range migration of the target through a scale transformation on the range frequency-slow time plane. Among them, the second-order KT (SoKT) can correct the second-order range migration of the target through a second-order scale transformation. However, the KT type method may have a trajectory segmentation effect when detecting maneuvering targets, and the detection performance is reduced; 2) Radon transform type methods: Through multi-dimensional search in the motion parameter space, the energy of the target echo signal in different distance units is accumulated, and the RM and DFM are corrected at the same time, the echo signal-to-noise ratio is improved, thereby improving the target detection performance, but this type of method has high computational complexity; 3) Non-parametric search processing type methods: Using the coupling relationship between echo signals, the signal energy accumulation is completed in the correlation function domain, and the target detection statistic related to the accumulated signal characteristics is constructed. The computational complexity is lower than that of the parameter search type method, but the detection probability is reduced.

[0004] In general, for the estimation and compensation of motion parameters of maneuvering targets, the existing long-term coherent accumulation methods involve high-dimensional optimization problems and have high computational complexity. The first type of method does not consider the essential reason for the KT trajectory segmentation effect in the maneuvering target detection scenario. It directly performs KT under the condition that the target Doppler frequency is ambiguous. The Doppler frequency exceeds the slow time sampling frequency, resulting in a decrease in KT detection performance and a large error in the estimation of target motion parameters. The second type of method does not consider the differences in the formation mechanism of range migration and Doppler migration. It directly estimates the optimal compensation functions such as speed and acceleration based on the minimum granularity criterion, resulting in a coupled estimation problem between the range and Doppler of the target motion parameters. Therefore, there is a lot of computational redundancy and the algorithm has low degree of freedom. Although the computational complexity of the third type of non-parametric search processing method is greatly reduced, the performance also drops sharply. Therefore, for the needs of maneuvering weak target detection, it is urgent to study an efficient long-term coherent accumulation detection method for maneuvering weak targets. On the one hand, its computational complexity can meet the actual system requirements, and on the other hand, its detection probability is comparable to the performance of the optimal detector. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides an efficient long-term coherent accumulation detection method for maneuvering weak targets.

[0006] In order to achieve the above object, the technical solution provided by the present invention is: an efficient long-term coherent accumulation detection method for maneuvering weak targets, comprising the following steps:

[0007] Step 1: Construct the slant range variable search space between the radar and the maneuvering target based on the radar parameters and the motion characteristics of the maneuvering target and motion parameter variable search space in: They represent the target radial distance variable search space, radial velocity variable search space and radial acceleration variable search space respectively; c 0,min and c 0,max Respectively represent the minimum and maximum values ​​of the target radial distance variable; c 1,min and c 1,max Respectively represent the minimum and maximum values ​​of the target radial velocity variable; c 2,min and c 2,max Respectively represent the minimum and maximum values ​​of the target radial acceleration variable;

[0008] Step 2: Variable search space of target motion parameters and Perform three-scale decomposition to obtain the target fuzzy motion parameter variable search space required for Doppler frequency defuzzification The search space of target coarse-scale baseband motion parameter variables determined by the minimum motion parameter scale required for target range motion correction The search space of target fine-scale motion parameter variables is determined by the minimum motion parameter scale required by Doppler motion correction. in: They represent the target fuzzy velocity variable search space and the fuzzy acceleration variable search space respectively; They represent the target coarse-scale baseband radial velocity variable search space and the coarse-scale baseband radial acceleration variable search space respectively; They represent the target fine-scale baseband radial velocity variable search space and the fine-scale baseband radial acceleration variable search space respectively.

[0009] 2.1. Three-scale decomposition formula of target motion parameter variables:

[0010]

[0011] in: is the target p-order motion parameter variable; is the target p-order fuzzy motion parameter variable; is the target p-order baseband motion parameter variable, is the Doppler frequency after deambiguation the remainder; is the target coarse-scale p-order baseband motion parameter variable, which is determined by the minimum scale of motion parameters required by range migration correction; is the target fine-scale p-order baseband motion parameter variable, which is determined by the minimum scale of motion parameters required by Doppler migration correction and is the value after range migration correction. The remainder.

[0012] Specifically, according to the radar parameters, in order to avoid the segmentation effect of the Keystone transform (KT) trajectory caused by Doppler ambiguity, the target p-order fuzzy motion parameter variable is defined as:

[0013]

[0014] in: represents the target p-order motion parameter fuzzy number; c p represents the true value of the target p-order motion parameter; c p,min represents the minimum value of the target p-order motion parameter variable search space; round(·) represents rounding operation; Δc p,a It is the minimum p-order motion parameter interval that ensures that the Doppler frequency is not ambiguous, that is, the Doppler frequency does not exceed [-PRF / 2,PRF / 2]:

[0015]

[0016] Where: 0≤β p ≤1, p=1,2 satisfies PRF is the radar pulse repetition frequency; λ is the wavelength of the radar transmit signal; T is the radar coherent integration time.

[0017] Depend on The value range of the baseband motion parameters meets the following conditions:

[0018]

[0019] Therefore, in order to avoid range migration caused by baseband motion parameters, the target coarse-scale p-order baseband motion parameter variable is defined as:

[0020]

[0021] Where: The target coarse-scale p-order motion parameter search interval Δ RM,p is the minimum interval on the distance unit to ensure that no p-order distance migration occurs:

[0022]

[0023] Where: 0≤α p ≤1, p=1,2 satisfies c is the speed of light; f s is the fast-time sampling frequency.

[0024] To avoid Doppler migration, the target fine-scale p-order baseband motion parameter variable is defined as:

[0025]

[0026] Where: Target fine-scale p-order baseband motion parameter search interval Δ DFM,p is the minimum interval on the Doppler unit to ensure that no p-order Doppler migration occurs:

[0027]

[0028] Where: f c is the carrier frequency.

[0029] 2.2. Based on the three-scale decomposition formula of target motion parameters, the target motion parameter variable search space is decomposed into three scales to obtain the target fuzzy scale motion parameter search space, the target coarse scale baseband motion parameter search space and the target fine scale baseband motion parameter search space.

[0030] The target fuzzy scale p-order motion parameter variable search space is defined as follows:

[0031]

[0032] The target coarse-scale p-order baseband motion parameter variable search space is defined as follows:

[0033]

[0034] The target fine-scale p-order baseband motion parameter variable search space is defined as follows:

[0035]

[0036] Step 3: For an array of fuzzy scale motion parameter variables Construct the fuzzy motion parameter compensation function in the target fuzzy scale motion parameter search space:

[0037]

[0038] Where: f k = kΔf, k = 1, ..., K represents the fast time domain frequency; Δf is the frequency interval; K is the number of frequency intervals; t m =mPRT, m=1,…, M represents slow time, PRT is the pulse repetition time; M is the number of pulses.

[0039] The Doppler frequency defuzzification of the echo signal is performed using the fuzzy scale motion parameter compensation function as shown below:

[0040]

[0041] Among them: echo signal A represents the complex amplitude of the target in the frequency domain; c0, c1 and c2 represent the radial distance, radial velocity and radial acceleration of the target at the initial moment respectively; c 1,b and c 2,b represent the target radial baseband velocity and radial baseband acceleration respectively.

[0042] Step 4: Compensate the Doppler ambiguity frequency for the signal y1(f k ,t m ) to perform Keystone transformation, that is, Eliminate linear distance migration of targets:

[0043]

[0044] Where: A2 = A1exp[-j4πc0 / λ]; t m′ ≈m′PRT is the slow time variable after Keystone transformation, m′=0,…,M-1; λ=c / f c is the wavelength of the radar signal.

[0045] Step 5: For an array of coarse-scale baseband motion parameter variables In the target coarse-scale baseband motion parameter search space, the echo signal y2(f k ,t m′ ) to perform modified Generalized Inverse Fourier Transform (mGIFT):

[0046]

[0047] Where: τ n =n / f s (n=1,…,K) represents fast time; f s is the fast time domain sampling frequency; is the target coarse-scale baseband motion parameter distance migration compensation function; is the target coarse-scale baseband motion parameter Doppler migration pre-compensation function; A3(τ n )=A2Kexp[jπf s (τ n -2c0 / c)] is the complex amplitude of the signal after mGIFT transformation; the scale factor

[0048] Step 6: For a set of p ≥ 2 fine-scale baseband motion parameter variables, an array is constructed. p is the order of motion parameters, For y3(τ n ,t m′ ) is subjected to a generalized Fourier transform (GFT) to obtain a temporary range-Doppler (RD) spectrum:

[0049]

[0050] in: represents an array of fine-scale baseband motion parameter variables multiplied by the scale factor κ; A4(τ n ,v d )=A3(τ n )Mexp[-j2πT(v d +c 1,b ) / λ]; is the target baseband speed variable; is based on Doppler migration compensation function; represents the fine-scale baseband acceleration variable transformed by the scale factor κ.

[0051] Step 7: Based on the temporary RD spectrum plane obtained in step 6, determine whether the target exists using the likelihood ratio test (LRT) method:

[0052]

[0053] Where: γ LRT is the threshold of the likelihood ratio test; H0 is the null hypothesis of LRT; H1 is the alternative hypothesis of LRT. If LRT is established, all radial motion parameters of the target are estimated: target radial distance estimate is the τ obtained by the likelihood ratio test n Estimated value; target radial velocity estimate Target radial acceleration estimate c is obtained by the likelihood ratio test 1,b The estimated value, and its relationship with the estimated value of the target radial velocity is: is the target radial baseband velocity estimate.

[0054] Step 8: Under the target fuzzy scale motion parameters and coarse scale baseband motion parameters given in steps 3 and 5, if the target fine scale baseband motion parameter search space in step 6 is not traversed Then return to step 6; if the target fine-scale baseband motion parameter search space has been traversed Then proceed to the next step.

[0055] Step 9: If the target coarse-scale baseband motion parameter variable search space in step 5 is not traversed Then return to step 5; if the target coarse-scale baseband motion parameter variable search space has been traversed Then proceed to the next step.

[0056] Step 10: If the target fuzzy scale motion parameter variable search space in step 3 is not traversed Then return to step 3; if the target fuzzy scale motion parameter variable search space has been traversed Then proceed to the next step.

[0057] Step 11: Based on the estimated blur scale motion parameters of all targets from step 1 to step 10 Target baseband motion speed Target coarse-scale baseband motion parameters Target fine-scale baseband motion parameters And target radial distance J represents the number of targets, and estimates the radial distances and all motion parameters of all targets. in: and They represent the radial distance estimation, fuzzy scale radial velocity estimation and fuzzy scale radial acceleration estimation of the j-th target respectively; and They represent the baseband radial velocity estimate, coarse-scale baseband radial acceleration estimate, and fine-scale baseband radial acceleration estimate of the j-th target, respectively; They represent the radial velocity estimate and radial acceleration estimate of the j-th target respectively.

[0058] Using the radial distance estimation and motion parameter estimation of all targets, the fuzzy Doppler frequency of the target is compensated, the Keystone transformation is performed, and then the remaining range migration and Doppler migration are corrected to estimate the target RD spectrum:

[0059]

[0060] in: represents the complex amplitude of the jth target after TS-KT-MFP algorithm; is the estimated Doppler frequency of the jth target, that is The asinc function is the discrete form of the sinc function. n )=sin(πBτ n ) / [K valid sin(πΔfτ n )]; B = K valid Δf is the signal bandwidth; Δf is the frequency unit spacing; K valid is the number of effective frequency units; PRF = 1 / PRT is the pulse repetition frequency; is the radial distance of the jth target.

[0061] The target motion parameters And RD spectrum is used for subsequent radar signal processing.

[0062] The beneficial effects of the present invention relative to the prior art are:

[0063] 1) Improve the algorithm's degree of freedom by decomposing the motion parameter variables into three scales;

[0064] 2) In the target fuzzy scale motion parameter variable search space, the fuzzy effect of the echo signal Doppler frequency is eliminated through the compensation function to avoid the trajectory segmentation effect of KT, thereby improving the accumulation performance of the algorithm;

[0065] 3) In the target coarse-scale motion parameter variable search space, the range migration is corrected by GIFT transformation to achieve complete range migration compensation without ignoring the range migration caused by the motion parameter residue;

[0066] 4) The LRT criterion is used to detect targets on the RD spectrum, and the cross terms between multiple targets at the same distance are suppressed by threshold limiting, thereby reducing false target detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 The present invention provides a flowchart of an efficient long-term coherent accumulation detection method for maneuvering weak targets according to an embodiment of the present invention. DETAILED DESCRIPTION

[0068] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0069] Reference Figure 1 This embodiment provides an efficient long-term coherent accumulation detection method for maneuvering weak targets, comprising the following steps:

[0070] Step 1: Construct the slant range variable search space between the radar and the maneuvering target based on the radar parameters and the motion characteristics of the maneuvering target and motion parameter variable search space in: They represent the target radial distance variable search space, radial velocity variable search space and radial acceleration variable search space respectively; c 0,min and c 0,max Respectively represent the minimum and maximum values ​​of the target radial distance variable; c 1,min and c 1,max Respectively represent the minimum and maximum values ​​of the target radial velocity variable; c 2,min and c 2,max Respectively represent the minimum and maximum values ​​of the target radial acceleration variable;

[0071] Step 2: Variable search space of target motion parameters and Perform three-scale decomposition to obtain the target fuzzy motion parameter variable search space required for Doppler frequency defuzzification The search space of target coarse-scale baseband motion parameter variables determined by the minimum motion parameter scale required for target range motion correction The search space of target fine-scale motion parameter variables is determined by the minimum motion parameter scale required by Doppler motion correction. in: They represent the target fuzzy velocity variable search space and the fuzzy acceleration variable search space respectively; They represent the target coarse-scale baseband radial velocity variable search space and the coarse-scale baseband radial acceleration variable search space respectively; They represent the target fine-scale baseband radial velocity variable search space and the fine-scale baseband radial acceleration variable search space respectively.

[0072] 2.1. Three-scale decomposition formula of target motion parameter variables:

[0073]

[0074] in: is the target p-order motion parameter variable; is the target p-order fuzzy motion parameter variable; is the target p-order baseband motion parameter variable, is the Doppler frequency after deambiguation the remainder; is the target coarse-scale p-order baseband motion parameter variable, which is determined by the minimum scale of motion parameters required by range migration correction; is the target fine-scale p-order baseband motion parameter variable, which is determined by the minimum scale of motion parameters required by Doppler migration correction and is the value after range migration correction. The remainder.

[0075] Specifically, according to the radar parameters, in order to avoid the KT trajectory segmentation effect caused by Doppler ambiguity, the target p-order fuzzy motion parameter variable is defined as:

[0076]

[0077] in: represents the target p-order motion parameter fuzzy number; c p represents the true value of the target p-order motion parameter; c p,min represents the minimum value of the target p-order motion parameter variable search space; round(·) represents rounding operation; Δc p,a It is the minimum p-order motion parameter interval that ensures that the Doppler frequency is not ambiguous, that is, the Doppler frequency does not exceed [-PRF / 2,PRF / 2]:

[0078]

[0079] Where: 0≤β p ≤1, p=1,2 satisfies PRF is the radar pulse repetition frequency; λ is the wavelength of the radar transmit signal; T is the radar coherent integration time.

[0080] Depend on The value range of the baseband motion parameters meets the following conditions:

[0081]

[0082] Therefore, in order to avoid range migration caused by baseband motion parameters, the target coarse-scale p-order baseband motion parameter variable is defined as:

[0083]

[0084] Where: The target coarse-scale p-order motion parameter search interval Δ RM,p is the minimum interval on the distance unit to ensure that no p-order distance migration occurs:

[0085]

[0086] Where: 0≤α p ≤1, p=1,2 satisfies c is the speed of light; f s is the fast-time sampling frequency.

[0087] To avoid Doppler migration, the target fine-scale p-order baseband motion parameter variable is defined as:

[0088]

[0089] Where: Target fine-scale p-order baseband motion parameter search interval Δ DFM,p is the minimum interval on the Doppler unit to ensure that no p-order Doppler migration occurs:

[0090]

[0091] Where: f c is the carrier frequency.

[0092] 2.2. Based on the three-scale decomposition formula of target motion parameters, the target motion parameter variable search space is decomposed into three scales to obtain the target fuzzy scale motion parameter search space, the target coarse scale baseband motion parameter search space and the target fine scale baseband motion parameter search space.

[0093] The target fuzzy scale p-order motion parameter variable search space is defined as follows:

[0094]

[0095] The target coarse-scale p-order baseband motion parameter variable search space is defined as follows:

[0096]

[0097] The target fine-scale p-order baseband motion parameter variable search space is defined as follows:

[0098]

[0099] Step 3: For an array of fuzzy scale motion parameter variables Construct the fuzzy motion parameter compensation function in the target fuzzy scale motion parameter search space:

[0100]

[0101] Where: f k = kΔf, k = 1, ..., K represents the fast time domain frequency; Δf is the frequency interval; K is the number of frequency intervals; t m =mPRT, m=1,…, M represents slow time, PRT is the pulse repetition time; M is the number of pulses.

[0102] The Doppler frequency defuzzification of the echo signal is performed using the fuzzy scale motion parameter compensation function as shown below:

[0103]

[0104] Among them: echo signal A represents the complex amplitude of the target in the frequency domain; c0, c1 and c2 represent the radial distance, radial velocity and radial acceleration of the target at the initial moment respectively; c 1,b and c 2,b represent the target radial baseband velocity and radial baseband acceleration respectively.

[0105] Step 4: Compensate the Doppler ambiguity frequency for the signal y1(f k ,t m ) to perform Keystone transformation, that is, Eliminate linear distance migration of targets:

[0106]

[0107] Where: A2 = A1exp[-j4πc0 / λ]; t m′ ≈m′PRT is the slow time variable after Keystone transformation, m′=0,…,M-1; λ=c / f c is the wavelength of the radar signal.

[0108] Step 5: For an array of coarse-scale baseband motion parameter variables In the target coarse-scale baseband motion parameter search space, the echo signal y2(fk ,t m′ ) to perform modified Generalized Inverse Fourier Transform (mGIFT):

[0109]

[0110] Where: τ n =n / f s (n=1,…,K) represents fast time; f s is the fast time domain sampling frequency; is the target coarse-scale baseband motion parameter distance migration compensation function; is the target coarse-scale baseband motion parameter Doppler migration pre-compensation function; A3(τ n )=A2Kexp[jπf s (τ n -2c0 / c)] is the complex amplitude of the signal after mGIFT transformation; the scale factor

[0111] Step 6: For a set of p ≥ 2 fine-scale baseband motion parameter variables, an array is constructed. p is the order of motion parameters, For y3(τ n ,t m′ ) is subjected to a generalized Fourier transform (GFT) to obtain a temporary range-Doppler (RD) spectrum:

[0112]

[0113] in: represents an array of fine-scale baseband motion parameter variables multiplied by the scale factor κ; A4(τ n ,v d )=A3(τ n )Mexp[-j2πT(v d +c 1,b ) / λ]; is the target baseband speed variable; is based on Doppler migration compensation function; represents the fine-scale baseband acceleration variable transformed by the scale factor κ.

[0114] Step 7: Based on the temporary RD spectrum plane obtained in step 6, determine whether the target exists using the likelihood ratio test (LRT) method:

[0115]

[0116] Where: γ LRT is the threshold of the likelihood ratio test; H0 is the null hypothesis of LRT; H1 is the alternative hypothesis of LRT. If LRT is established, all radial motion parameters of the target are estimated: target radial distance estimate is the τ obtained by the likelihood ratio test n Estimated value; target radial velocity estimate Target radial acceleration estimate c is obtained by the likelihood ratio test 1,b The estimated value, and its relationship with the estimated value of the target radial velocity is: is the target radial baseband velocity estimate.

[0117] Step 8: Under the target fuzzy scale motion parameters and coarse scale baseband motion parameters given in steps 3 and 5, if the target fine scale baseband motion parameter search space in step 6 is not traversed Then return to step 6; if the target fine-scale baseband motion parameter search space has been traversed Then proceed to the next step.

[0118] Step 9: If the target coarse-scale baseband motion parameter variable search space in step 5 is not traversed Then return to step 5; if the target coarse-scale baseband motion parameter variable search space has been traversed Then proceed to the next step.

[0119] Step 10: If the target fuzzy scale motion parameter variable search space in step 3 is not traversed Then return to step 3; if the target fuzzy scale motion parameter variable search space has been traversed Then proceed to the next step.

[0120] Step 11: Based on the estimated blur scale motion parameters of all targets from step 1 to step 10 Target baseband motion speed Target coarse-scale baseband motion parameters Target fine-scale baseband motion parameters And target radial distance J represents the number of targets, and estimates the radial distances and all motion parameters of all targets. in: and They represent the radial distance estimation, fuzzy scale radial velocity estimation and fuzzy scale radial acceleration estimation of the j-th target respectively; and They represent the baseband radial velocity estimate, coarse-scale baseband radial acceleration estimate, and fine-scale baseband radial acceleration estimate of the j-th target, respectively; They represent the radial velocity estimate and radial acceleration estimate of the j-th target respectively.

[0121] Using the radial distance estimation and motion parameter estimation of all targets, the fuzzy Doppler frequency of the target is compensated, the Keystone transformation is performed, and then the remaining range migration and Doppler migration are corrected to estimate the target RD spectrum:

[0122]

[0123] in: represents the complex amplitude of the jth target after TS-KT-MFP algorithm; is the estimated Doppler frequency of the jth target, that is The asinc function is the discrete form of the sinc function. n )=sin(πBτ n ) / [K valid sin(πΔfτ n )]; B = K valid Δf is the signal bandwidth; Δf is the frequency unit spacing; K valid is the number of effective frequency units; PRF = 1 / PRT is the pulse repetition frequency; is the radial distance of the jth target.

[0124] The target motion parameters And RD spectrum is used for subsequent radar signal processing.

[0125] The embodiments of the present invention have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. An efficient long-term coherent integration detection method for maneuvering weak targets, characterized in that: The following steps are involved: Step 1: Based on the radar parameters and the motion characteristics of the maneuvering target, construct the slant range variable search space between the radar and the maneuvering target and motion parameter variable search space in: They represent the target radial distance variable search space, radial velocity variable search space and radial acceleration variable search space respectively; c 0,min and c 0,max Respectively represent the minimum and maximum values ​​of the target radial distance variable; c 1,min and c 1,max Respectively represent the minimum and maximum values ​​of the target radial velocity variable; c 2,min and c 2,max Respectively represent the minimum and maximum values ​​of the target radial acceleration variable; Step 2: Variable search space of target motion parameters and Perform three-scale decomposition to obtain the target fuzzy motion parameter variable search space required for Doppler frequency defuzzification The search space of target coarse-scale baseband motion parameter variables determined by the minimum motion parameter scale required for target range motion correction The search space of target fine-scale motion parameter variables is determined by the minimum motion parameter scale required by Doppler motion correction. in: They represent the target fuzzy velocity variable search space and the fuzzy acceleration variable search space respectively; They represent the target coarse-scale baseband radial velocity variable search space and the coarse-scale baseband radial acceleration variable search space respectively; They represent the target fine-scale baseband radial velocity variable search space and the fine-scale baseband radial acceleration variable search space respectively; Step 3: For an array of fuzzy scale motion parameter variables Constructing fuzzy motion parameter compensation function in target fuzzy scale motion parameter search space Where: f k = kΔf, k = 1, ..., K represents the fast time domain frequency; Δf is the frequency interval; K is the number of frequency intervals; t m =mPRT, m=1,…, M represents slow time, PRT is the pulse repetition time; M is the number of pulses. use The Doppler frequency of the echo signal is deambiguated to obtain the Doppler frequency unambiguous signal y1(f k ,t m ); Step 4: Compensate the Doppler ambiguity frequency for the signal y1(f k ,t m ) to perform Keystone transformation, that is, Eliminate linear distance migration of targets: Where: A2 = A1exp[-j4πc0 / λ]; A1 is y1(f k ,t m ) amplitude; t m′ ≈m′PRT is the slow time variable after Keystone transformation, m′=0,…,M-1; c is the speed of light; f c is the carrier frequency; λ = c / f c is the wavelength of the radar transmission signal; c0, c 1,b and c 2,b They represent the target radial distance, radial baseband velocity and radial baseband acceleration at the initial moment respectively. Step 5: For an array of coarse-scale baseband motion parameter variables In the target coarse-scale baseband motion parameter search space, the echo signal y2(f k ,t m′ ) is used to perform an improved generalized inverse Fourier transform, and we get y3(τ n ,t m′ ); Step 6: For a set of p ≥ 2 fine-scale baseband motion parameter variables, an array is constructed. p represents the order of motion parameters. For y3(τ n ,t m′ ) performs a generalized Fourier transform to obtain the temporary range-Doppler spectrum: in: represents an array of fine-scale baseband motion parameter variables multiplied by the scale factor κ; A4(τ n ,v d )=A3(τ n )Mexp[-j2πT(v d +c 1,b ) / λ]; is the target baseband speed variable; is based on Doppler migration compensation function; represents the fine-scale baseband acceleration variable transformed by the scale factor κ; Step 7: Based on the temporary range-Doppler spectrum plane obtained in step 6, determine whether there is a target according to the likelihood ratio test. If there is a target, estimate all the motion parameters of the target. Step 8: Under the target fuzzy scale motion parameters and coarse scale baseband motion parameters given in steps 3 and 5, if the target fine scale baseband motion parameter search space in step 6 is not traversed Then return to step 6; if the target fine-scale baseband motion parameter search space has been traversed Then proceed to the next step; Step 9: If the target coarse-scale baseband motion parameter variable search space in step 5 is not traversed Then return to step 5; if the target coarse-scale baseband motion parameter variable search space has been traversed Then proceed to the next step; Step 10: If the target fuzzy scale motion parameter variable search space in step 3 is not traversed Then return to step 3; if the target fuzzy scale motion parameter variable search space has been traversed Then proceed to the next step; Step 11: Based on the estimated blur scale motion parameters of all targets from step 1 to step 10 Target baseband motion speed Target coarse-scale baseband motion parameters Target fine-scale baseband motion parameters And target radial distance J represents the number of targets, and estimates the radial distances and all motion parameters of all targets. in: and They represent the radial distance estimation, fuzzy scale radial velocity estimation and fuzzy scale radial acceleration estimation of the j-th target respectively; and They represent the baseband radial velocity estimate, coarse-scale baseband radial acceleration estimate, and fine-scale baseband radial acceleration estimate of the j-th target, respectively; They represent the radial velocity estimate and radial acceleration estimate of the j-th target respectively; Using the radial distance estimation and motion parameter estimation of all targets, the fuzzy Doppler frequency of the target is compensated, the Keystone transformation is performed, and then the remaining range migration and Doppler migration are corrected to estimate the target RD spectrum: in: represents the complex amplitude of the jth target after TS-KT-MFP algorithm; is the estimated Doppler frequency of the jth target, that is The asinc function is the discrete form of the sinc function. n )=sin(πBτ n ) / [K valid sin(πΔfτ n )]; B = K valid Δf is the signal bandwidth; Δf is the frequency unit spacing; K valid is the number of effective frequency units; PRF = 1 / PRT is the pulse repetition frequency; is the radial distance of the jth target; The motion parameters of the maneuvering target and range-Doppler spectrum for subsequent radar signal processing.

2. The method for detecting a highly efficient maneuvering weak target by long-term coherent integration as claimed in claim 1, characterized in that: In step 2, the target motion parameter three-scale decomposition method includes: Step 2.1: Three-scale decomposition formula of target motion parameter variables: in: is the target p-order motion parameter variable; is the target p-order fuzzy motion parameter variable; is the target p-order baseband motion parameter variable, is the Doppler frequency after deambiguation the remainder; is the target coarse-scale p-order baseband motion parameter variable, which is determined by the minimum scale of motion parameters required for range migration correction; is the target fine-scale p-order baseband motion parameter variable, which is determined by the minimum scale of motion parameters required by Doppler migration correction and is the value after range migration correction. the remainder; The target p-order fuzzy motion parameter variable is defined as: in: represents the target p-order motion parameter fuzzy number; c p represents the true value of the target p-order motion parameter; c p,min represents the minimum value of the search space of the target p-order motion parameter variable; round(·) represents the rounding operation; Δc p,a It is the minimum p-order motion parameter interval that ensures that the Doppler frequency is not ambiguous, that is, the Doppler frequency does not exceed [-PRF / 2,PRF / 2]: Where: 0≤β p ≤1, p=1,2 satisfies PRF is the radar pulse repetition frequency; λ is the wavelength of the radar transmission signal; T is the radar coherent integration time; Depend on The value range of the baseband motion parameters meets the following conditions: The target coarse-scale p-order baseband motion parameter variable is defined as: Where: The target coarse-scale p-order motion parameter search interval Δ RM,p is the minimum interval on the distance unit to ensure that no p-order distance migration occurs: Where: 0≤α p ≤1, p=1,2 satisfies c is the speed of light; f s is the fast-time sampling frequency. The target fine-scale p-order baseband motion parameter variable is defined as: Where: Target fine-scale p-order baseband motion parameter search interval Δ DFM,p is the minimum interval on the Doppler unit to ensure that no p-order Doppler migration occurs: Where: f c is the carrier frequency; Step 2.2, based on the three-scale decomposition formula of the target motion parameter, the target motion parameter variable search space is decomposed into three scales to obtain the target fuzzy scale motion parameter search space, the target coarse scale baseband motion parameter search space and the target fine scale baseband motion parameter search space; The target fuzzy scale p-order motion parameter variable search space is defined as follows: The target coarse-scale p-order baseband motion parameter variable search space is defined as follows: The target fine-scale p-order baseband motion parameter variable search space is defined as follows:

3. The efficient long-term coherent integration detection method for maneuvering weak targets as claimed in claim 1, characterized in that: The fuzzy motion parameter compensation formula in step 3 is: Among them: echo signal A represents the complex amplitude of the target in the frequency domain; c0, c1 and c2 represent the radial distance, radial velocity and radial acceleration of the target at the initial moment respectively; fuzzy motion parameter compensation function f k = kΔf, k = 1, ..., K represents the fast time domain frequency; Δf is the frequency interval; K is the number of frequency intervals; t m =mPRT, m=1,…, M represents slow time, PRT is the pulse repetition time; M is the number of pulses.

4. The efficient long-term coherent integration detection method for maneuvering weak targets as claimed in claim 1, characterized in that: The mGIFT transformation formula in step 5 is: For an array of coarse-scale motion parameter variables In the target coarse-scale motion parameter search space, the echo signal y2(f k ,t m′ ) to perform mGIFT transformation: Where: τ n =n / f s (n=1,…,K) represents fast time; f s is the fast time sampling frequency; is the target coarse-scale baseband motion parameter distance migration compensation function; is the target coarse-scale baseband motion parameter Doppler migration pre-compensation function; A3(τ n )=A2Kexp[jπf s (τ n -2c0 / c)] is the complex amplitude of the signal after mGIFT transformation; the scale factor 5. The efficient long-term coherent integration detection method for maneuvering weak targets as claimed in claim 1, characterized in that: The likelihood ratio test formula in step 7 is as follows: Where: γ LRT is the threshold of the likelihood ratio test; H0 is the null hypothesis of LRT; H1 is the alternative hypothesis of LRT. If LRT is established, all radial motion parameters of the target are estimated: target radial distance estimate is the τ obtained by the likelihood ratio test n Estimated value; target radial velocity estimate Target radial acceleration estimate c is obtained by the likelihood ratio test 1,b The estimated value, and its relationship with the estimated value of the target radial velocity is: is the target radial baseband velocity estimate.

Images