A long-term coherent integration method and system for uniformly accelerated maneuvering targets

Abstract

The present invention belongs to the field of radar signal processing technology, and in particular, relates to a long-term coherent accumulation method and system for a uniformly accelerated maneuvering target. The method comprises: performing pulse compression processing on a target echo signal to obtain a pulse compressed signal; performing a second-order Keystone transform on the pulse compressed signal to obtain a second-order Keystone transformed signal; determining a search distance, a search baseband unambiguous speed, and a search fuzzy speed multiple of the search parameter according to a preset search parameter range, traversing the combination of the search distance, the search baseband unambiguous speed, and the search fuzzy speed multiple of each search parameter, performing two-dimensional search and coherent accumulation on the second-order Keystone transformed signal in a search distance-search speed domain, and obtaining coherent accumulation results of all targets; performing target detection on the coherent accumulation result, and estimating motion parameters of the detected target.

Description

Technical Field

[0001] The invention belongs to the technical field of radar signal processing and radar maneuvering target detection, and in particular, relates to a long-term coherent accumulation method and system for uniformly accelerated maneuvering targets. Background Art

[0002] With the continuous development of aerospace technology and the increasing maturity of modern stealth technology, how to accurately and effectively detect weak maneuvering targets has become a difficult problem in the field of radar signal processing. Under the premise of not changing the radar hardware and basic parameters as much as possible, long-term coherent integration is a technology that can effectively improve the detection of weak targets by radar. Uniformly accelerated motion is a common motion form of maneuvering targets. When a maneuvering target in a uniformly accelerated motion state is observed by the radar for a long time, its velocity will cause the echo to have first-order range migration, and its acceleration will cause the echo to have second-order range migration and Doppler diffusion. When the radar echo of the maneuvering target is coherently integrated, the main lobe will be widened and the accumulation peak will decrease, which will affect the radar's detection performance of the maneuvering target. Therefore, before the radar coherent accumulation detection, the range migration and Doppler migration must be eliminated.

[0003] For solving the first-order range migration caused by the speed of maneuvering targets, the more typical algorithms are Keystone transform and Radon-Fourier transform. Among them, Keystone transform generally uses interpolation operation to achieve the scale transformation of the two-dimensional data plane, thereby effectively correcting the first-order range migration (reference [1]: RPPerry, et al., "SAR imaging of moving targets," IEEE Transactions on Aerospace and Electronic Systems, vol. 35, no. 1, pp. 188-200, 1999). Radon-Fourier transform (reference [2]: J.Xu, et al., "Radon-Fourier Transform for Radar Target Detection, I: Generalized Doppler FilterBank," IEEE Transactions on Aerospace and Electronic Systems, vol. 47, no. 2, pp. 1186-1202, 2011) performs coherent accumulation of maneuvering target energy through a two-dimensional joint search of distance and speed. These two methods can only correct the first-order range movement. When the maneuvering target performs uniform acceleration motion and exhibits second-order range movement and Doppler movement, the coherent integration effect will be significantly deteriorated.

[0004] In order to solve the problem of second-order range migration and Doppler spread in the echo of uniformly accelerated maneuvering targets, Xu Jia et al. proposed a generalized Radon-Fourier transform (reference [3]: J.Xu, et al., “Radar Maneuvering Target Motion Estimation Based on Generalized Radon-Fourier Transform,” IEEE Transactions on Signal Processing, vol. 60, no. 12, pp. 6190-6201, 2012). It is a generalized definition of the existing Radon-Fourier transform. It performs a coherent accumulation of the energy of the maneuvering target through a three-dimensional joint search of distance, velocity, and acceleration. In addition, algorithms such as RFRFT, RLVD, and RLCT have been proposed to solve the above problems (Reference [4]: ​​XLChen, et al., “Maneuvering Target Detection via Radon-FractionalFourier Transform-Based Long-Time Coherent Integration,” IEEE Transactions on Signal Processing, vol. 62, no. 4, pp. 939-953, 2014; Reference [5]: XL Li, et al., “Coherent Integration for Maneuvering Target Detection Based on Radon-Lv's Distribution,” IEEE Signal Processing Letters, vol. 22, no. 9, pp. 1467-1471, 2015; Reference [6]: XLChen, et al., “Detection of a Low Observable Sea-Surface Target With Micromotion via the Radon-Linear Canonical Transform,” IEEE Geoscience and Remote Sensing Letters, vol.11, no.7, pp.1225-1229, 2014.). However, existing methods all jointly search in three dimensions: distance, speed, and acceleration, and eventually accumulate the energy of maneuvering targets in three-dimensional or even four-dimensional space. Then, target detection also needs to be performed in three-dimensional or even four-dimensional space, which results in high computational complexity and affects the real-time performance of radar signal processing.Existing methods suffer from the problem of high computational complexity. Summary of the invention

[0005] In order to solve the above-mentioned defects in the prior art, the present invention proposes a long-time coherent integration method for uniformly accelerated maneuvering targets, specifically relates to a convenient long-time coherent integration method with low computational complexity, which can reduce the computational complexity. The method eliminates the range migration and Doppler movement in the radar echo of the uniformly accelerated maneuvering target, realizes the coherent accumulation of target energy and the estimation of target motion parameters.

[0006] The present invention provides a long-term coherent integration method for a uniformly accelerated maneuvering target, the method comprising:

[0007] The radar uses a linear frequency modulation signal as the transmitting signal. The radar receiver receives the target echo signal and performs pulse compression processing on the target echo signal to obtain a pulse compressed signal.

[0008] Perform a second-order Keystone transform on the pulse compressed signal to obtain a second-order Keystone transformed signal;

[0009] According to the range of the pre-set search parameters, the search distance, search baseband unambiguous speed and search ambiguity speed multiple of the search parameters are determined, and the combination of the search distance, search baseband unambiguous speed and search ambiguity speed multiple of each search parameter is traversed, and a two-dimensional search and coherent accumulation are performed in the search distance-search speed domain for the signal after the second-order Keystone transformation to obtain the coherent accumulation result;

[0010] Target detection is performed on the coherent accumulation results, and motion parameters of the detected targets are estimated.

[0011] As one of the improvements of the above technical solution, the radar uses a linear frequency modulation signal as a transmitting signal, and the radar receiver receives the target echo signal, performs pulse compression processing on the target echo signal, and obtains a pulse compressed signal; the specific implementation process is:

[0012] The radar uses a linear frequency modulation signal as the transmitting signal, and the radar receiver receives the target echo signal s raw (t r ,t m ):

[0013]

[0014] Among them, A r is the amplitude of the echo signal; t r For fast time; represents a unit imaginary number; c is the speed of light; exp(·) represents an exponential function with the natural logarithm e as the base; f cis the center frequency of the transmitted signal; T p is the duration of the transmitted signal; K r is the modulation frequency of the transmitted linear frequency modulation signal; t m is the slow time, t m =mT r ; Where m = 0, 1, 2, ... M; T r is the pulse repetition interval, M represents the total number of pulses accumulated coherently;

[0015] R t (t m ) is the target and the radar at t m The instantaneous distance of a moment:

[0016]

[0017] Where R0 is the initial distance of the target relative to the radar; a r represents the acceleration of the target; v r Express the radial velocity of the target relative to the radar:

[0018] v r =v b +M amb ·v prf (3)

[0019] Among them, v b To search for the baseband unambiguous velocity, v b ∈[-v prf / 2,v prf / 2]; M amb v is the search fuzzy speed multiple; prf =λ×PRF / 2 is the fuzzy speed; where PRF is the pulse repetition frequency of the radar, PRF=1 / T r ;λ is the wavelength of the transmitted signal, λ=c / f c ;

[0020] The target echo signal s raw (t r ,t m ) to perform pulse compression processing and obtain the pulse compressed signal s rc (t r ,t m ):

[0021]

[0022] Among them, A rc is the signal amplitude after pulse compression; B r is the transmit signal bandwidth, B r =K r T p .

[0023] As one of the improvements of the above technical solution, the second-order Keystone transform is performed on the pulse compressed signal to obtain a second-order Keystone transformed signal; the specific implementation process is:

[0024] The pulse compressed signal s rc (t r ,t m ) along the fast time t r Perform Fourier transform to obtain the fast time frequency-slow time domain signal S rm (f r ,t m ):

[0025]

[0026] Among them, f r is the fast time t r The corresponding fast time frequency, A rm is the signal amplitude;

[0027] For S rm (f r ,t m ) m Perform the following variable substitutions:

[0028]

[0029] Among them, t n is the new slow time variable after variable substitution;

[0030] After variable substitution, the signal becomes:

[0031]

[0032] For S rm (f r ,t n ) along the fast time t r Do the inverse Fourier transform and set r = ct r / 2, and the signal s after the second-order Keystone transformation is obtained sokt (r,t n ):

[0033]

[0034] Among them, v r1 =v hb +M amb v prf ;in, v hb ∈[-v prf / 4,v prf / 4].

[0035] As one of the improvements of the above technical solution, the search distance, search baseband unambiguous speed and search fuzzy speed multiple of the search parameter are determined according to the preset search parameter range, and the combination of the search distance, search baseband unambiguous speed and search fuzzy speed multiple of each search parameter is traversed. According to the obtained second-order Keystone transformed signal, two-dimensional search and coherent accumulation are performed in the search distance-search speed domain to obtain the coherent accumulation result; the specific implementation process is:

[0036] Assume the search distance r sc , search baseband unambiguous speed v hb_sc and search fuzzy speed multiple m amb_sc ;

[0037] Set the search distance r sc The search range is [r min ,r max ], the corresponding search step is Δr, then the search distance is Then we get:

[0038] r sc =n r Δr+r min (9)

[0039] Among them, n r =0,1,2,…,N r -1,n r is the number of the search distance; r min is the search distance r sc The minimum value of r max is the search distance r sc The maximum value of

[0040] Set the search speed to v sc :

[0041] v sc =v hb_sc +m amb_sc v prf ;

[0042] It should be noted that v sc The corresponding is v r1 , rather than the target's actual velocity v r ;

[0043] Among them, v hb_sc To search for baseband unambiguous velocity, the search range is [-v prf / 4,v prf / 4], the corresponding search step is Δv, then the number of baseband unambiguous velocity searches is N v , Then we get:

[0044]

[0045] m amb_sc is the search fuzzy speed multiple, and its search range is [M amb_min ,M amb_max ], the corresponding search step is 1, and then we get:

[0046] m amb_sc =M amb_min ,M amb_min +1,M amb_min +2,…,M amb_max -1,M amb_max ;

[0047] Among them, M amb_min is the minimum value of the search fuzzy speed multiple; M amb_max The maximum value of the search fuzzy speed multiplier;

[0048] According to the search distance r set above sc and search speed v sc , extract the signal s after the second-order Keystone transformation sokt (r,t n ) sokt (r sc +v sc t n ,t n ), and perform discrete chirp-Fourier transform, and then estimate the quadratic term coefficient as a according to the peak position of the transformed result. r_est ;

[0049] According to a r_est and v hb_sc Construct the phase compensation equation H v_a (t n ):

[0050]

[0051] The constructed phase compensation equation H v_a (t n ) into the following Radon-Fourier transform algorithm formula, in the search distance-search speed domain (r sc ,v sc ) to perform coherent accumulation and obtain the coherent accumulation result R rv (r sc ,v sc ):

[0052]

[0053] Among them, T CPI T is the coherent accumulation time; CPI =MT r .

[0054] As one of the improvements of the above technical solution, the target detection is performed on the coherent accumulation result, and the motion parameter estimation is performed on the detected target; the specific process is:

[0055] The coherent accumulation result R rv (r sc ,v sc ) to perform target detection;

[0056] If the coherent accumulation result R rv (r sc ,v sc ) is less than the preset threshold value, it is determined that no target is detected;

[0057] If the coherent accumulation result R rv (r sc ,v sc ) is greater than or equal to a preset threshold value, it is determined that a target is detected, and motion parameters of the detected target are estimated;

[0058] According to R rv (r sc ,v sc ) The peak point position corresponds to r sc , get the distance R between the target and the radar 0_est ;

[0059] According to R rv (r sc ,v sc ) The peak point position corresponds to v sc The corresponding v hb_sc and m amb_sc , calculate the radial velocity v of the target relative to the radar r_est :

[0060] v r_est =2v hb_sc +m amb_sc ·v prf (13)

[0061] According to s sokt (r sc +v sc t n ,t n) is used to perform discrete chirp-Fourier transform on the peak position to obtain the acceleration a of the target r_est .

[0062] The present invention also provides a long-term coherent accumulation system for a uniformly accelerated maneuvering target, the system comprising:

[0063] A pulse compression module is used to perform pulse compression processing on the echo signal to obtain a pulse compressed signal;

[0064] A second-order Keystone transform module, used to perform a second-order Keystone transform on the pulse compressed signal to obtain a second-order Keystone transformed signal;

[0065] A coherent accumulation module is used to determine the search distance, search baseband unambiguous speed and search fuzzy speed multiple of the search parameter according to the preset search parameter range, traverse the combination of the search distance, search baseband unambiguous speed and search fuzzy speed multiple of each search parameter, and perform two-dimensional search and coherent accumulation in the search distance-search speed domain according to the obtained second-order Keystone transformed signal to obtain a coherent accumulation result;

[0066] The target detection module is used to perform target detection on the coherent accumulation results and estimate the motion parameters of the detected targets.

[0067] As one of the improvements of the above technical solution, the specific implementation process of the pulse compression module is:

[0068] The radar uses a linear frequency modulation signal as the transmitting signal, and the radar receiver receives the target echo signal s raw (t r ,t m ):

[0069]

[0070] Among them, A r is the amplitude of the echo signal; t r For fast time; represents a unit imaginary number; c is the speed of light; exp(·) represents an exponential function with the natural logarithm e as the base; f c is the center frequency of the transmitted signal; T p is the duration of the transmitted signal; K r is the modulation frequency of the transmitted linear frequency modulation signal; t m is the slow time, t m =mT r ; Where m = 0, 1, 2, ... M; T r is the pulse repetition interval, M represents the total number of pulses accumulated coherently;

[0071] R t (t m ) is the target and the radar at t m The instantaneous distance of a moment:

[0072]

[0073] Where R0 is the initial distance of the target relative to the radar; a r represents the acceleration of the target; v r Express the radial velocity of the target relative to the radar:

[0074] v r =v b +M amb ·v prf (16)

[0075] Among them, v b To search for the baseband unambiguous velocity, v b ∈[-v prf / 2,v prf / 2]; M amb v is the search fuzzy speed multiple; prf =λ×PRF / 2 is the fuzzy speed; where PRF is the pulse repetition frequency of the radar, PRF=1 / T r ;λ is the wavelength of the transmitted signal, λ=c / f c ;

[0076] For the echo signal s raw (t r ,t m ) to perform pulse compression processing and obtain the pulse compressed signal s rc (t r ,t m ):

[0077]

[0078] Among them, A rc is the signal amplitude after pulse compression; B r is the transmit signal bandwidth, B r =K r T p .

[0079] As one of the improvements of the above technical solution, the specific implementation process of the second-order Keystone transformation module is:

[0080] The pulse compressed signal s rc (t r ,t m ) along the fast time t rPerform Fourier transform to obtain the fast time frequency-slow time domain signal S rm (f r ,t m ):

[0081]

[0082] Among them, f r is the fast time t r The corresponding fast time frequency, A rm is the signal amplitude.

[0083] For S rm (f r ,t m ) m Perform the following variable substitutions:

[0084]

[0085] Among them, t n is the new slow time variable after variable substitution;

[0086] After variable substitution, the signal becomes:

[0087]

[0088] For S rm (f r ,t n ) along the fast time t r Do the inverse Fourier transform and set r = ct r / 2, and the signal s after the second-order Keystone transformation is obtained sokt (r,t n ):

[0089]

[0090] Among them, v r1 =v hb +M amb v prf ;in, v hb ∈[-v prf / 4,v prf / 4].

[0091] As one of the improvements of the above technical solution, the specific implementation process of the coherent accumulation module is:

[0092] Assume the search distance r sc , search baseband unambiguous speed v hb_sc and search fuzzy speed multiple m amb_sc ;

[0093] Set the search distance r sc The search range is [r min ,r max ], the corresponding search step is Δr, then the search distance is Then we get:

[0094] r sc =n r Δr+r min (twenty two)

[0095] Among them, n r =0,1,2,…,N r -1, is the number of the search distance; r min is the search distance r sc The minimum value of r max is the search distance r sc The maximum value of

[0096] Set the search speed to v sc :

[0097] v sc =v hb_sc +m amb_sc v prf ;

[0098] It should be noted that v sc The corresponding is v r1 , rather than the target's actual velocity v r ;

[0099] Among them, v hb_sc To search for baseband unambiguous velocity, the search range is [-v prf / 4,v prf / 4], the corresponding search step is Δv, then the number of baseband unambiguous velocity searches is N v , Then we get:

[0100]

[0101] m amb_sc is the search fuzzy speed multiple, and its search range is [M amb_min ,M amb_max ], the corresponding search step is 1, and then we get:

[0102] m amb_sc =M amb_min ,M amb_min +1,M amb_min +2,…,M amb_max -1,M amb_max ;

[0103] Among them, M amb_min is the minimum value of the search fuzzy speed multiple; M amb_max The maximum value of the search fuzzy speed multiplier;

[0104] According to the search distance r set above sc and search speed v sc , extract the signal s after the second-order Keystone transformation sokt (r,t n ) sokt (r sc +v sc t n ,t n ), and perform discrete chirp-Fourier transform, and then estimate the quadratic term coefficient as a according to the peak position of the transformed result. r_est ;

[0105] According to a r_est and v hb_sc Construct the phase compensation equation H v_a (t n ):

[0106]

[0107] The constructed phase compensation equation H v_a (t n ) into the following Radon-Fourier transform algorithm formula, in the search distance-search speed domain (r sc ,v sc ) to perform coherent accumulation and obtain the coherent accumulation result R rv (r sc ,v sc ):

[0108]

[0109] Among them, T CPI T is the coherent accumulation time; CPI =MT r .

[0110] As one of the improvements of the above technical solution, the specific implementation process of the target detection module is as follows:

[0111] The coherent accumulation result R rv (r sc ,v sc ) to perform target detection;

[0112] If the coherent accumulation result R rv (r sc ,v sc) is less than the preset threshold value, it is determined that no target is detected;

[0113] If the coherent accumulation result R rv (r sc ,v sc ) is greater than or equal to a preset threshold value, it is determined that a target is detected, and motion parameters of the detected target are estimated;

[0114] According to R rv (r sc ,v sc ) The peak point position corresponds to r sc , get the distance R between the target and the radar 0_est ;

[0115] According to R rv (r sc ,v sc ) The peak point position corresponds to v sc The corresponding v hb_sc and m amb_sc , calculate the radial velocity v of the target relative to the radar r_est :

[0116] v r_est =2v hb_sc +m amb_sc ·v prf (26)

[0117] According to s sokt (r sc +v sc t n ,t n ) is used to perform discrete chirp-Fourier transform on the peak position to obtain the acceleration a of the target r_est .

[0118] The beneficial effects of the present invention compared with the prior art are:

[0119] The method of the present invention simultaneously utilizes the amplitude and phase information in the radar echo of the maneuvering target to perform long-term coherent accumulation, which can effectively improve the radar echo signal-to-noise ratio, thereby improving the radar's detection performance of the target; in addition, the method of the present invention only performs a two-dimensional search of distance and speed, accumulates the target echo energy in the distance-speed two-dimensional space, and target detection only needs to be performed in the distance-speed two-dimensional space, with low computational complexity, convenient for real-time processing of radar signals, conducive to engineering implementation, and has promotion and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0120] Figure 1 It is a flow chart of a long-time coherent accumulation method for a uniformly accelerated maneuvering target provided by the present invention;

[0121] Figure 2 This is a schematic diagram of the result after the echo pulse received by the radar is compressed;

[0122] Figure 3 It is a schematic diagram of the coherent accumulation result in the long-term coherent accumulation method for a uniformly accelerated maneuvering target provided by the present invention;

[0123] Figure 4 A schematic diagram of the result of discrete chirp-Fourier transform of the signal extracted when searching for the distance and speed corresponding to the target in the long-term coherent accumulation method for a uniformly accelerated maneuvering target provided by the present invention;

[0124] Figure 5 It is a schematic diagram of the existing moving target detection (MTD) result;

[0125] Figure 6 It is a schematic diagram of the Radon-Fourier transform result of the prior art;

[0126] Figure 7 It is a schematic diagram of the generalized Radon-Fourier transform result of the prior art. DETAILED DESCRIPTION

[0127] The present invention will now be further described with reference to the accompanying drawings and examples.

[0128] The present invention provides a long-term coherent integration method and system for a uniformly accelerated maneuvering target, and specifically relates to a long-term coherent integration method and system for a uniformly accelerated maneuvering target radar echo based on second-order Keystone and Radon-Fourier transform;

[0129] The method includes: performing pulse compression on the radar received target echo signal, then using the second-order Keystone transform to correct the second-order range migration, then using the distance and speed joint search to achieve the coherent accumulation of target energy, and finally performing target detection and parameter estimation on the coherent accumulation result. The present invention solves the range migration and Doppler movement problems that appear in the echo when the radar observes the uniformly accelerated target for a long time, effectively improves the radar echo signal-to-noise ratio, and thus improves the detection performance of the target. In addition, the present invention only performs a two-dimensional search of distance and speed, accumulates the target echo energy in the distance-speed two-dimensional space, and target detection only needs to be performed in the distance-speed two-dimensional space. The computational complexity is low, which is convenient for real-time processing of radar signals, is conducive to engineering implementation, and has promotion and application value.

[0130] like Figure 1 As shown, the method specifically includes:

[0131] Step S1) Target echo distance pulse compression

[0132] The radar uses a linear frequency modulation signal as the transmitting signal. The radar receiver receives the target echo signal and performs pulse compression processing on the target echo signal to obtain a pulse compressed signal.

[0133] Specifically, the radar uses a linear frequency modulation signal as the transmitting signal, and the radar receiver receives the target echo signal s raw (t r ,t m ):

[0134]

[0135] Among them, A r is the amplitude of the echo signal; t r For fast time; represents a unit imaginary number; c is the speed of light; exp(·) represents an exponential function with the natural logarithm e as the base; f c is the center frequency of the transmitted signal; T p K is the duration of the transmitted signal; r is the modulation frequency of the transmitted linear frequency modulation signal; t m is the slow time, t m =mT r ; Where m = 0, 1, 2, .. M; T r is the pulse repetition interval, M represents the total number of pulses accumulated coherently;

[0136] R t (t m ) is the target and the radar at t m The instantaneous distance of a moment:

[0137]

[0138] Where R0 is the initial distance of the target relative to the radar; a r represents the acceleration of the target; v r Express the radial velocity of the target relative to the radar:

[0139] v r =v b +M amb ·v prf (29)

[0140] Among them, v b To search for the baseband unambiguous velocity, v b ∈[-v prf / 2,v prf / 2]; M amb v is the search fuzzy speed multiple; prf=λ×PRF / 2 is the fuzzy speed; where PRF is the pulse repetition frequency of the radar, PRF=1 / T r ;λ=c / f c , is the wavelength of the transmitted signal;

[0141] For the echo signal s raw (t r ,t m ) to perform pulse compression processing to obtain the pulse compressed signal:

[0142]

[0143] Among them, A rc is the signal amplitude after pulse compression; B r is the transmit signal bandwidth, B r =K r T p ;

[0144] Step S2) Second-order Keystone transform

[0145] The pulse compressed signal s rc (t r ,t m ) along the fast time t r Perform Fourier transform to obtain the fast time frequency-slow time domain signal S rm (f r ,t m ):

[0146]

[0147] Among them, f r is the fast time t r The corresponding fast time frequency, A rm is the signal amplitude.

[0148] For S rm (f r ,t m ) m Perform the following variable substitutions:

[0149]

[0150] Among them, t n is the new slow time variable after variable substitution;

[0151] After variable substitution, the signal becomes:

[0152]

[0153] For S rm (f r ,tn ) along the fast time t r Do the inverse Fourier transform and set r = ct r / 2, and the signal s after the second-order Keystone transformation is obtained sokt (r,t n ):

[0154]

[0155] In the above formula, v r1 =v hb +M amb v prf ;in, v hb ∈[-v prf / 4,v prf / 4].

[0156] After the second-order Keystone transform, the baseband unambiguous velocity v hb It has become half of its original size.

[0157] Step S3) Distance-speed two-dimensional search to achieve coherent accumulation

[0158] According to the range of the pre-set search parameters, the search distance, the search baseband unambiguous speed and the search ambiguity speed multiple of the search parameters are determined, and the combination of the search distance, the search baseband unambiguous speed and the search ambiguity speed multiple of each search parameter is traversed. According to the obtained second-order Keystone transformed signal, two-dimensional search and coherent accumulation are performed in the search distance-search speed domain to obtain the coherent accumulation result;

[0159] Specifically, assuming the search distance r sc , search baseband unambiguous speed v hb_sc and search fuzzy speed multiple m amb_sc ;

[0160] Set the search distance r sc The search range is [r min ,r max ], the corresponding search step is Δr, then the search distance is Then we get:

[0161] r sc =n r Δr+r min (35)

[0162] Among them, n r =0,1,2,…,N r -1,n r is the number of the search distance; r min is the search distance rsc The minimum value of r max is the search distance r sc The maximum value of

[0163] Set the search speed to v sc :

[0164] v sc =v hb_sc +m amb_sc v prf ;

[0165] It should be noted that v sc The corresponding is v r1 , rather than the target's actual velocity v r ;

[0166] Among them, v hb_sc To search for baseband unambiguous velocity, the search range is [-v prf / 4,v prf / 4], the corresponding search step is Δv, then the number of baseband unambiguous velocity searches is N v , Then we get:

[0167]

[0168] m amb_sc is the search fuzzy speed multiple, and its search range is [M amb_min ,M amb_max ], the corresponding search step is 1, and then we get:

[0169] m amb_sc =M amb_min ,M amb_min +1,M amb_min +2,…,M amb_max -1,M amb_max ;

[0170] Among them, M amb_min is the minimum value of the search fuzzy speed multiple; M amb_max The maximum value of the search fuzzy speed multiplier;

[0171] According to the search distance r set above sc and search speed v sc , extract the signal s after the second-order Keystone transformation sokt (r,t n ) sokt (r sc +v sc t n ,t n), and perform discrete chirp-Fourier transform, and then estimate the quadratic term coefficient as a according to the peak position of the transformed result. r_est ;

[0172] According to a r_est and v hb_sc Construct the phase compensation equation H v_a (t n ):

[0173]

[0174] The constructed phase compensation equation H v_a (t n ) is brought into the following Radon-Fourier transform algorithm formula, in the search distance-search speed domain (r sc ,v sc ) to perform coherent accumulation and obtain the coherent accumulation result R rv (r sc ,v sc ):

[0175]

[0176] Among them, T CPI T is the coherent accumulation time; CPI =MT r .

[0177] Step S4) Target detection and motion parameter estimation

[0178] Target detection is performed on the coherent accumulation results, and motion parameters of the detected targets are estimated.

[0179] Specifically, the coherent accumulation result R rv (r sc ,v sc ) to perform target detection;

[0180] If the coherent accumulation result R rv (r sc ,v sc ) is less than the preset threshold value, it is determined that no target is detected;

[0181] If the coherent accumulation result R rv (r sc ,v sc ) is greater than or equal to a preset threshold value, it is determined that a target is detected, and motion parameters of the detected target are estimated;

[0182] According to R rv (r sc ,v sc) The peak point position corresponds to r sc , get the distance R between the target and the radar 0_est ;

[0183] According to R rv (r sc ,v sc ) The peak point position corresponds to v sc The corresponding v hb_sc and m amb_sc , calculate the radial velocity v of the target relative to the radar r_est :

[0184] v r_est =2v hb_sc +m amb_sc ·v prf (39)

[0185] According to s sokt (r sc +v sc t n ,t n ) is used to perform discrete chirp-Fourier transform on the peak position to obtain the acceleration a of the target r_est .

[0186] The present invention also provides a long-term coherent accumulation system for a uniformly accelerated maneuvering target, the system comprising:

[0187] A pulse compression module is used to perform pulse compression processing on the echo signal to obtain a pulse compressed signal;

[0188] Specifically, the radar uses a linear frequency modulation signal as the transmitting signal, and the radar receiver receives the target echo signal s raw (t r ,t m ):

[0189]

[0190] Among them, A r is the amplitude of the echo signal; t r For fast time; represents a unit imaginary number; c is the speed of light; exp(·) represents an exponential function with the natural logarithm e as the base; f c is the center frequency of the transmitted signal; T p is the duration of the transmitted signal; K r is the modulation frequency of the transmitted linear frequency modulation signal; t m is the slow time, t m =mT r ; Where m = 0, 1, 2, ... M; T ris the pulse repetition interval, M represents the total number of pulses accumulated coherently;

[0191] R t (t m ) is the target and the radar at t m The instantaneous distance of a moment:

[0192]

[0193] Where R0 is the initial distance of the target relative to the radar; a r represents the acceleration of the target; v r Express the radial velocity of the target relative to the radar:

[0194] v r =v b +M amb ·v prf (42)

[0195] Among them, v b To search for the baseband unambiguous velocity, v b ∈[-v prf / 2,v prf / 2]; M amb v is the search fuzzy speed multiple; prf =λ×PRF / 2 is the fuzzy speed; where PRF is the pulse repetition frequency of the radar, PRF=1 / T r ;λ is the wavelength of the transmitted signal, λ=c / f c ;

[0196] For the echo signal s raw (t r ,t m ) to perform pulse compression processing to obtain the pulse compressed signal:

[0197]

[0198] Among them, A rc is the signal amplitude after pulse compression; B r is the transmit signal bandwidth, B r =K r T p ;

[0199] A second-order Keystone transform module, used to perform a second-order Keystone transform on the pulse compressed signal to obtain a second-order Keystone transformed signal;

[0200] The pulse compressed signal s rc (t r ,t m ) along the fast time t rPerform Fourier transform to obtain the fast time frequency-slow time domain signal S rm (f r ,t m ):

[0201]

[0202] Among them, f r is the fast time t r The corresponding distance frequency, A rm is the signal amplitude.

[0203] For S rm (f r ,t m ) m Perform the following variable substitutions:

[0204]

[0205] Among them, t n is the new slow time variable after variable substitution;

[0206] After variable substitution, the signal becomes:

[0207]

[0208] For S rm (f r ,t n ) along the fast time t r Do the inverse Fourier transform and set r = ct r / 2, and the signal s after the second-order Keystone transformation is obtained sokt (r,t n ):

[0209]

[0210] In the above formula, v r1 =v hb +M amb v prf ;in, v hb ∈[-v prf / 4,v prf / 4].

[0211] A coherent accumulation module is used to determine the search distance, search baseband unambiguous speed and search fuzzy speed multiple of the search parameter according to the preset search parameter range, traverse the combination of the search distance, search baseband unambiguous speed and search fuzzy speed multiple of each search parameter, and perform two-dimensional search and coherent accumulation in the search distance-search speed domain according to the obtained second-order Keystone transformed signal to obtain a coherent accumulation result;

[0212] Specifically, assuming the search distance r sc , search baseband unambiguous speed v hb_sc and search fuzzy speed multiple m amb_sc ;

[0213] Set the search distance r sc The search range is [r min ,r max ], the corresponding search step is Δr, then the search distance is Then we get:

[0214] r sc =n r Δr+r min (48)

[0215] Among them, n r =0,1,2,…,N r -1,n r is the number of the search distance; r min is the search distance r sc The minimum value of r max is the search distance r sc The maximum value of

[0216] Set the search speed to v sc :

[0217] v sc =v hb_sc +m amb_sc v prf ;

[0218] It should be noted that v sc The corresponding is v r1 , rather than the target's actual velocity v r ;

[0219] Among them, v hb_sc To search for baseband unambiguous velocity, the search range is [-v prf / 4,v prf / 4], the corresponding search step is Δv, then the number of baseband unambiguous velocity searches is N v , Then we get:

[0220]

[0221] m amb_sc is the search fuzzy speed multiple, and its search range is [M amb_min ,M amb_max ], the corresponding search step is 1, and then we get:

[0222] m amb_sc =M amb_min ,M amb_min +1,M amb_min +2,…,M amb_max -1,M amb_max ;

[0223] Among them, M amb_min is the minimum value of the search fuzzy speed multiple; M amb_max The maximum value of the search fuzzy speed multiplier;

[0224] According to the search distance r set above sc and search speed v sc , extract the signal s after the second-order Keystone transform sokt (r,t n ) sokt (r sc +v sc t n ,t n ), and perform discrete chirp-Fourier transform, and then estimate the quadratic term coefficient as a according to the peak position of the transformed result. r_est ;

[0225] According to a r_est and v hb_sc Construct the phase compensation equation H v_a (t n ):

[0226]

[0227] The constructed phase compensation equation H v_a (t n ) into the following Radon-Fourier transform algorithm formula, in the search distance-search speed domain (r sc ,v sc ) to perform coherent accumulation and obtain the coherent accumulation result R rv (r sc ,v sc ):

[0228]

[0229] Among them, TCPI T is the coherent accumulation time; CPI =MT r .

[0230] The target detection module is used to perform target detection on the coherent accumulation results and estimate the motion parameters of the detected targets.

[0231] Specifically, the coherent accumulation result R rv (r sc ,v sc ) to perform target detection;

[0232] If the coherent accumulation result R rv (r sc ,v sc ) is less than the preset threshold value, it is determined that no target is detected;

[0233] If the coherent accumulation result R rv (r sc ,v sc ) is greater than or equal to a preset threshold value, it is determined that a target is detected, and motion parameters of the detected target are estimated;

[0234] According to R rv (r sc ,v sc ) The peak point position corresponds to r sc , get the distance R between the target and the radar 0_est ;

[0235] According to R rv (r sc ,v sc ) The peak point position corresponds to v sc The corresponding v hb_sc and m amb_sc , calculate the radial velocity v of the target relative to the radar r_est :

[0236] v r_est =2v hb_sc +m amb_sc ·v prf (52)

[0237] According to s sokt (r sc +v sc t n ,t n ) is used to perform discrete chirp-Fourier transform on the peak position to obtain the acceleration a of the target r_est .

[0238] like Figure 2-Figure 7As shown, the method of the present invention is combined with simulation experiments to verify the long-term coherent integration method for the radar echo of a uniformly accelerated maneuvering target provided by the present invention.

[0239] Radar system transmits signal center frequency f c =1.25GHz, transmission signal bandwidth B r =40MHz, transmission time T p =5us, pulse repetition frequency PRF=1000, coherent accumulation pulse number M=1000, coherent accumulation time T CPI = 1s. The fuzzy speed v corresponding to this system prf =120m / s. The initial distance of the simulated target relative to the radar is R0=100km, and the radial velocity is v r =340m / s, radial acceleration is a r =30m / s 2 The baseband unambiguous velocity corresponding to this target is v b =-20m / s, fuzzy speed multiple M amb =3.

[0240] Figure 2 This is the result of echo pulse compression received by the radar. The signal-to-noise ratio after pulse compression is 3dB, which shows that the target signal envelope has undergone range migration. Figure 3 This is the coherent accumulation result of this method. It can be seen that the target signal is accumulated into a peak. Figure 4 This is the result of discrete chirp-Fourier transform of the signal extracted when this method searches for distance and speed corresponding to the target. Figure 3 The peak point position can be used to estimate the distance R of the target 0_est =99.99km, speed v r_est =340m / s; Figure 4 The peak position can be used to estimate the target acceleration a r_est =29.98m / s 2 .

[0241] To illustrate the effectiveness of this method, Figure 5 , Figure 6 , Figure 7 The coherent accumulation results of three existing typical methods are shown. Figure 5 and Figure 6 They are the results of moving target detection (MTD) and Radon-Fourier transform, and it can be seen that due to the influence of target echo distance and Doppler migration, the coherent accumulation performance of conventional moving target detection (MTD) and Radon-Fourier transform algorithms is significantly reduced, and no target accumulation peak can appear in either figure. Figure 7Show that the generalized Radon-Fourier transform is useful in searching for an acceleration value of 30 m / s 2 From the distance-velocity section obtained when , we can see that the target signal accumulates into a peak, and the amplitude is comparable to that of this method, but the computational complexity of this algorithm is significantly higher than that of this method.

[0242] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention is described in detail with reference to the embodiments, it should be understood by those skilled in the art that any modification or equivalent replacement of the technical solutions of the present invention does not depart from the spirit and scope of the technical solutions of the present invention and should be included in the scope of the claims of the present invention.

Claims

1. A long-term coherent integration method for a uniformly accelerated maneuvering target, characterized in that: The method includes: The radar uses a linear frequency modulation signal as the transmitting signal. The radar receiver receives the target echo signal and performs pulse compression processing on the echo signal to obtain a pulse compressed signal. Perform a second-order Keystone transform on the pulse compressed signal to obtain a second-order Keystone transformed signal; According to the range of the pre-set search parameters, the search distance, the search baseband unambiguous speed and the search ambiguity speed multiple of the search parameters are determined, and the combination of the search distance, the search baseband unambiguous speed and the search ambiguity speed multiple of each search parameter is traversed, and the signal after the second-order Keystone transformation is subjected to two-dimensional search and coherent accumulation in the search distance-search speed domain to obtain the coherent accumulation result; Perform target detection on the coherent accumulation results and estimate motion parameters of the detected targets; The radar uses a linear frequency modulation signal as a transmitting signal, and the radar receiver receives the target echo signal, performs pulse compression processing on the echo signal, and obtains a pulse compressed signal; the specific implementation process is: The radar uses a linear frequency modulation signal as the transmitting signal, and the radar receiver receives the target echo signal s raw (t r ,t m ): Among them, A r is the amplitude of the echo signal; t r For fast time; represents a unit imaginary number; c is the speed of light; exp(·) represents an exponential function with the natural logarithm e as the base; f c is the center frequency of the transmitted signal; T p is the duration of the transmitted signal; K r is the modulation frequency of the transmitted linear frequency modulation signal; t m is the slow time, t m =mT r ; Where m = 0, 1, 2, ... M; T r is the pulse repetition interval, M represents the total number of pulses accumulated coherently; R t (t m ) is the target and the radar at t m The instantaneous distance of a moment: Where R0 is the initial distance of the target relative to the radar; a r represents the acceleration of the target; v r Express the radial velocity of the target relative to the radar: v r =v b +M amb ·v prf (3) Among them, v b To search for the baseband unambiguous velocity, v b ∈[-v prf / 2,v prf / 2]; M amb v is the search fuzzy speed multiple; prf =λ×PRF / 2 is the fuzzy speed; where PRF is the pulse repetition frequency of the radar, PRF=1 / T r ;λ is the wavelength of the transmitted signal, λ=c / f c ; The target echo signal s raw (t r ,t m ) to perform pulse compression processing and obtain the pulse compressed signal s rc (t r ,t m ): Among them, A rc is the signal amplitude after pulse compression; B r is the transmit signal bandwidth, B r =K r T p ; The second-order Keystone transform is performed on the pulse compressed signal to obtain a second-order Keystone transformed signal; the specific implementation process is: The pulse compressed signal s rc (t r ,t m ) along the fast time t r Perform Fourier transform to obtain the fast time frequency-slow time domain signal S rm (f r ,t m ): Among them, f r is the fast time t r The corresponding fast time frequency, A rm is the signal amplitude; For S rm (f r ,t m ) m Perform the following variable substitutions: Among them, t n is the new slow time variable after variable substitution; After variable substitution, the signal becomes: For S rm (f r ,t n ) along the fast time t r Do the inverse Fourier transform and set r = ct r / 2, and the signal s after the second-order Keystone transformation is obtained sokt (r,t n ): In the above formula, v r1 =v hb +M amb v prf ;in, v hb ∈[-v prf / 4,v prf / 4]; The method determines the search distance, search baseband unambiguous speed and search fuzzy speed multiple of the search parameters according to the range of the preset search parameters, traverses the combination of the search distance, search baseband unambiguous speed and search fuzzy speed multiple of each search parameter, and performs two-dimensional search and coherent accumulation in the search distance-search speed domain according to the obtained second-order Keystone transformed signal to obtain a coherent accumulation result; the specific implementation process is as follows: Assume the search distance r sc , search baseband unambiguous speed v hb_sc and search fuzzy speed multiple m amb_sc ; Set the search distance r sc The search range is [r min ,r max ], the corresponding search step is Δr, then the search distance is Then we get: r sc =n r Δr+r min (9) Among them, n r =0,1,2,…,N r -1 is the number of the search distance; r min is the search distance r sc The minimum value of r max is the search distance r sc The maximum value of Set the search speed to v sc : v sc =v hb_sc +m amb_sc v prf ; It should be noted that v sc The corresponding is v r1 , rather than the target's actual velocity v r ; Among them, v hb_sc To search for baseband unambiguous velocity, the search range is [-v prf / 4,v prf / 4], the corresponding search step is Δv, then the number of baseband unambiguous velocity searches is N v , Then we get: m amb_sc is the search fuzzy speed multiple, and its search range is [M amb_min ,M amb_max ], the corresponding search step is 1, and then we get: m amb_sc =M amb_min ,M amb_min +1,M amb_min +2,…,M amb_max -1,M amb_max ; Among them, M amb_min is the minimum value of the search fuzzy speed multiple; M amb_max The maximum value of the search fuzzy speed multiplier; According to the search distance r set above sc and search speed v sc , extract the signal s after the second-order Keystone transform sokt (r,t n ) sokt (r sc +v sc t n ,t n ), and perform discrete chirp-Fourier transform, and then estimate the quadratic term coefficient as a according to the peak position of the transformed result. r_est ; According to a r_est and v hb_sc Construct the phase compensation equation H v_a (t n ): The constructed phase compensation equation H v_a (t n ) into the following Radon-Fourier transform algorithm formula, in the search distance-search speed domain (r sc ,v sc ) to perform coherent accumulation and obtain the coherent accumulation result R rv (r sc ,v sc ): Among them, T CPI T is the coherent accumulation time; CPI =MT r .

2. The long-term coherent integration method for a uniformly accelerated maneuvering target according to claim 1, characterized in that: The target detection is performed on the coherent accumulation result, and the motion parameter estimation is performed on the detected target; the specific process is: The coherent accumulation result R rv (r sc ,v sc ) to perform target detection; If the coherent accumulation result R rv (r sc ,v sc ) is less than the preset threshold value, it is determined that no target is detected; If the coherent accumulation result R rv (r sc ,v sc ) is greater than or equal to a preset threshold value, it is determined that a target is detected, and motion parameters of the detected target are estimated; According to R rv (r sc ,v sc ) The peak point position corresponds to r sc , get the distance R between the target and the radar 0_est ; According to R rv (r sc ,v sc ) The peak point position corresponds to v sc The corresponding v hb_sc and m amb_sc , calculate the radial velocity v of the target relative to the radar r_est : v r_est =2v hb_sc +m amb_sc ·v prf (13) According to s sokt (r sc +v sc t n ,t n ) is used to perform discrete chirp-Fourier transform on the peak position to obtain the acceleration a of the target r_est .

3. A long-term coherent integration system for uniformly accelerated maneuvering targets, characterized in that: The system includes: A pulse compression module is used to perform pulse compression processing on the echo signal to obtain a pulse compressed signal; A second-order Keystone transform module, used to perform a second-order Keystone transform on the pulse compressed signal to obtain a second-order Keystone transformed signal; A coherent accumulation module is used to determine the search distance, search baseband unambiguous speed and search fuzzy speed multiple of the search parameter according to the preset search parameter range, traverse the combination of the search distance, search baseband unambiguous speed and search fuzzy speed multiple of each search parameter, and perform two-dimensional search and coherent accumulation in the search distance-search speed domain according to the obtained second-order Keystone transformed signal to obtain a coherent accumulation result; A target detection module is used to perform target detection on the coherent accumulation results and estimate motion parameters of the detected targets; The specific implementation process of the pulse compression module is: The radar uses a linear frequency modulation signal as the transmitting signal, and the radar receiver receives the target echo signal s raw (t r ,t m ): Among them, A r is the amplitude of the echo signal; t r For fast time; represents a unit imaginary number; c is the speed of light; exp(·) represents an exponential function with the natural logarithm e as the base; f c is the center frequency of the transmitted signal; T p is the duration of the transmitted signal; K r is the modulation frequency of the transmitted linear frequency modulation signal; t m is the slow time, t m =mT r ; Where m = 0, 1, 2, ... M; T r is the pulse repetition interval, M represents the total number of pulses accumulated coherently; R t (t m ) is the target and the radar at t m The instantaneous distance of a moment: Where R0 is the initial distance of the target relative to the radar; a r represents the acceleration of the target; v r Express the radial velocity of the target relative to the radar: v r =v b +M amb ·v prf (16) Among them, v b To search for the baseband unambiguous velocity, v b ∈[-v prf / 2,v prf / 2]; M amb v is the search fuzzy speed multiple; prf =λ×PRF / 2 is the fuzzy speed; where PRF is the pulse repetition frequency of the radar, PRF=1 / T r ;λ is the wavelength of the transmitted signal, λ=c / f c ; For the echo signal s raw (t r ,t m ) to perform pulse compression processing and obtain the pulse compressed signal s rc (t r ,t m ): Among them, A rc is the signal amplitude after pulse compression; B r is the transmit signal bandwidth, B r =K r T p ; The specific implementation process of the second-order Keystone transformation module is: After pulse compression, the signal s rc (t r ,t m ) along the fast time t r Perform Fourier transform to obtain the fast time frequency-slow time domain signal S rm (f r ,t m ): Among them, f r is the fast time t r The corresponding fast time frequency, A rm is the signal amplitude; For S rm (f r ,t m ) m Perform the following variable substitutions: Among them, t n is the new slow time variable after variable substitution; After variable substitution, the signal becomes: For S rm (f r ,t n ) along the fast time t r Do the inverse Fourier transform and set r = ct r / 2, and the signal s after the second-order Keystone transformation is obtained sokt (r,t n ): Among them, v r1 =v hb +M amb v prf ;in, v hb ∈[-v prf / 4,v prf / 4]; The specific implementation process of the coherent accumulation module is: Assume the search distance r sc , search baseband unambiguous speed v hb_sc and search fuzzy speed multiple m amb_sc ; Set the search distance r sc The search range is [r min ,r max ], the corresponding search step is Δr, then the search distance is Then we get: r sc =n r Δr+r min (22) Among them, n r =0,1,2,…,N r -1,n r is the number of the search distance; r min is the search distance r sc The minimum value of r max is the search distance r sc The maximum value of Set the search speed to v sc : v sc =v hb_sc +m amb_sc v prf ; It should be noted that v sc The corresponding is v r1 , rather than the target's actual velocity v r ; Among them, v hb_sc To search for baseband unambiguous velocity, the search range is [-v prf / 4,v prf / 4], the corresponding search step is Δv, then the number of baseband unambiguous velocity searches is N v , Then we get: m amb_sc is the search fuzzy speed multiple, and its search range is [M amb_min ,M amb_max ], the corresponding search step is 1, and then we get: m amb_sc =M amb_min ,M amb_min +1,M amb_min +2,…,M amb_max -1,M amb_max ; Among them, M amb_min is the minimum value of the search fuzzy speed multiple; M amb_max The maximum value of the search fuzzy speed multiplier; According to the search distance r set above sc and search speed v sc , extract the signal s after the second-order Keystone transform sokt (r,t n ) sokt (r sc +v sc t n ,t n ), and perform discrete chirp-Fourier transform, and then estimate the quadratic term coefficient as a according to the peak position of the transformed result. r_est ; According to a r_est and v hb_sc Construct the phase compensation equation H v_a (t n ): The constructed phase compensation equation H v_a (t n ) into the following Radon-Fourier transform algorithm formula, in the search distance-search speed domain (r sc ,v sc ) to perform coherent accumulation and obtain the coherent accumulation result R rv (r sc ,v sc ): Among them, T CPI T is the coherent accumulation time; CPI =MT r .

4. The long-term coherent integration system for uniformly accelerated maneuvering targets according to claim 3, characterized in that: The specific implementation process of the target detection module is as follows: The coherent accumulation result R rv (r sc ,v sc ) to perform target detection; If the coherent accumulation result R rv (r sc ,v sc ) is less than the preset threshold value, it is determined that no target is detected; If the coherent accumulation result R rv (r sc ,v sc ) is greater than or equal to a preset threshold value, it is determined that a target is detected, and motion parameters of the detected target are estimated; According to R rv (r sc ,v sc ) The peak point position corresponds to r sc , get the distance R between the target and the radar 0_est ; According to R rv (r sc ,v sc ) The peak point position corresponds to v sc The corresponding v hb_sc and m amb_sc , calculate the radial velocity v of the target relative to the radar r_est : v r_est =2v hb_sc +m amb_sc ·v prf (26) According to s sokt (r sc +v sc t n ,t n ) is used to perform discrete chirp-Fourier transform on the peak position to obtain the acceleration a of the target r_est .

Images