WO2022022239A1 - 一种基于检测前估计的雷达目标检测方法 - Google Patents
一种基于检测前估计的雷达目标检测方法 Download PDFInfo
- Publication number
- WO2022022239A1 WO2022022239A1 PCT/CN2021/104677 CN2021104677W WO2022022239A1 WO 2022022239 A1 WO2022022239 A1 WO 2022022239A1 CN 2021104677 W CN2021104677 W CN 2021104677W WO 2022022239 A1 WO2022022239 A1 WO 2022022239A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- detection
- target
- trace
- estimated
- target detection
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/50—Systems of measurement based on relative movement of target
- G01S13/505—Systems of measurement based on relative movement of target using Doppler effect for determining closest range to a target or corresponding time, e.g. miss-distance indicator
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/06—Systems determining position data of a target
- G01S13/08—Systems for measuring distance only
- G01S13/10—Systems for measuring distance only using transmission of interrupted, pulse modulated waves
- G01S13/26—Systems for measuring distance only using transmission of interrupted, pulse modulated waves wherein the transmitted pulses use a frequency- or phase-modulated carrier wave
- G01S13/28—Systems for measuring distance only using transmission of interrupted, pulse modulated waves wherein the transmitted pulses use a frequency- or phase-modulated carrier wave with time compression of received pulses
- G01S13/282—Systems for measuring distance only using transmission of interrupted, pulse modulated waves wherein the transmitted pulses use a frequency- or phase-modulated carrier wave with time compression of received pulses using a frequency modulated carrier wave
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/418—Theoretical aspects
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/50—Systems of measurement based on relative movement of target
- G01S13/58—Velocity or trajectory determination systems; Sense-of-movement determination systems
- G01S13/581—Velocity or trajectory determination systems; Sense-of-movement determination systems using transmission of interrupted pulse modulated waves and based upon the Doppler effect resulting from movement of targets
- G01S13/582—Velocity or trajectory determination systems; Sense-of-movement determination systems using transmission of interrupted pulse modulated waves and based upon the Doppler effect resulting from movement of targets adapted for simultaneous range and velocity measurements
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/50—Systems of measurement based on relative movement of target
- G01S13/58—Velocity or trajectory determination systems; Sense-of-movement determination systems
- G01S13/583—Velocity or trajectory determination systems; Sense-of-movement determination systems using transmission of continuous unmodulated waves, amplitude-, frequency-, or phase-modulated waves and based upon the Doppler effect resulting from movement of targets
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/66—Radar-tracking systems; Analogous systems
- G01S13/72—Radar-tracking systems; Analogous systems for two-dimensional [2D] tracking, e.g. combination of angle and range tracking, track-while-scan radar
- G01S13/723—Radar-tracking systems; Analogous systems for two-dimensional [2D] tracking, e.g. combination of angle and range tracking, track-while-scan radar by using numerical data
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
Definitions
- the invention relates to the field of radar signal processing, in particular to a radar target detection method based on pre-detection estimation.
- the method based on sparse restoration provides a feasible idea to solve the above problems.
- the reconstruction and parameter estimation of the target can be achieved by solving x.
- x is sparse, and solving x can be boiled down to a sparse recovery problem.
- object detection based on sparse reconstruction methods still faces three challenges: low-complexity reconstruction under ultra-high dimensional conditions, accurate reconstruction under off-grid conditions, and constant false alarm detection based on reconstruction results.
- the present invention provides a radar target detection method based on pre-detection estimation.
- the generalized maximum likelihood ratio detector is used for target detection, and the target detection results and parameters are output.
- the present invention After establishing a dimensionality reduction observation model based on the point trace parameter estimation results, the present invention further utilizes the target sparse prior information, and adopts the sparse recovery algorithm for target reconstruction and detection, which solves the three problems faced by the target detection method based on sparse recovery.
- the key issues are dimensionality reduction in ultra-high dimensional conditions, accurate reconstruction in off-grid conditions, and constant false alarm target detection based on nonlinear reconstruction results.
- the present invention completes target parameter estimation before detection, making it possible to "match" signal processing, greatly reducing signal processing loss and improving target detection performance.
- the simulation and measured data verification results show that, compared with the conventional method, the target detection performance of the present invention is improved by more than 1.5dB.
- FIG. 1 is a flowchart of a radar target detection method based on pre-detection estimation of the present invention
- FIG. 2 is a schematic diagram of the estimation accuracy of dot trace parameters provided by an embodiment of the present invention.
- FIG. 3 is a schematic diagram of statistical characteristics of reconstruction output noise based on a dimensionality reduction model provided by an embodiment of the present invention
- FIG. 4 is a schematic diagram showing the comparison of the target detection performance of four detectors under simulation conditions provided by an embodiment of the present invention
- FIG. 5 is a schematic diagram of a range Doppler spectrum of measured data provided by an embodiment of the present invention.
- FIG. 6 is a schematic diagram showing the performance comparison between the method of the present invention and the conventional method under the condition of measured data.
- the present invention is suitable for the pulse system radar using chirp, and the target detection processing flow is shown in Fig. 1, including the following steps:
- Step 1) perform conventional pulse compression and coherent accumulation on the received baseband data to obtain a range Doppler spectrum, and perform pre-detection based on the range Doppler spectrum to obtain a point of interest trace;
- the baseband data can include fast time, slow time and spatial domain data, and the present invention is suitable for processing three domain data. Since the air domain data model is similar to the slow time domain, the present invention takes the fast time and slow time domains as examples to introduce. Assuming that in a coherent processing interval, the number of distance cells is L and the number of pulses is K, the received signal can be expressed as a matrix Ignoring the effect of distance Doppler walking, its elements can be expressed as:
- y l, k are the l-th distance unit and the k-th pulse echo signal, 0 ⁇ k ⁇ K-1, 0 ⁇ l ⁇ L-1.
- P is the number of targets
- T s is the sampling time interval
- ⁇ p is the p-th target amplitude
- r p is the target distance
- T pul is the pulse duration
- TI is the pulse interval
- ⁇ is the frequency modulation slope
- c the speed of light.
- f d, p 2 ⁇ p / ⁇ is the Doppler frequency shift
- ⁇ p is the target velocity
- ⁇ is the carrier wavelength.
- n l, k are additive white Gaussian noise.
- j represents an imaginary number
- rect represents a rectangle function.
- Pulse compression in the fast time domain and coherent accumulation in the slow time domain can obtain the range Doppler spectrum. Then, based on the range Doppler spectrum, the conventional adaptive constant false alarm detection method is used to pre-detect the target and obtain the point of interest trace.
- the distance and Doppler frequency corresponding to the range unit and Doppler unit are respectively used and Express.
- Step 2) Estimate the point trace distance and Doppler parameter obtained by pre-detection, and obtain the estimated value and
- the received signal can be approximated as:
- y represents the received signal of a coherent processing interval, which is obtained by arranging Y into a vector in columns
- ⁇ is the dimensionality reduction target vector
- n is an additive white Gaussian noise vector.
- the Doppler frequency is The Doppler domain steering vector corresponding to the point trace of , represents the distance of The fast-time domain steering vector corresponding to the point trace of . and They are:
- the superscript T means transpose.
- ⁇ is still sparse, and ⁇ can be reconstructed based on sparse restoration, assuming that the reconstruction result is means, then, based on target detection can be achieved.
- the dot traces are not located in integer units, i.e. and There is a deviation from the real value of the point trace, and the reconstruction directly based on formula (2) faces the off-grid problem.
- the present invention is first based on and True parameters for point traces and Estimation is performed, followed by object reconstruction and detection.
- the received signal can be expressed as:
- w d represents the normalized window function in the slow time domain
- w ⁇ ,i represents the normalized window function in the fast time domain of the i-th trace
- w ⁇ ,i [q w,0 ,...,qw ,l , ...q w, L-1 ] T
- w ⁇ ,i [q w,0 ,...,qw ,l , ...q w, L-1 ] T
- ⁇ c (t) is the continuous form of the window function in time T pul .
- Equation (14) weights the data in the fast time and slow time domains.
- the i-th trace can be decoupled from other traces in the echo, that is, the i-th trace and other traces can be decoupled. There is approximately no mutual influence.
- g( ⁇ ) can be further approximated as:
- ⁇ i [r i , f i ] T , ri and f i represent the distance and Doppler frequency of the i-th trace respectively
- a i represents the steering vector of the i-th trace
- a ⁇ i means to remove the i-th column of the matrix A
- ⁇ ⁇ i means to remove the i-th element of the vector I.
- minimizing g( ⁇ ) must be such that minimum.
- the real parameters corresponding to the i-th trace can be obtained by minimizing u i ( ⁇ i ) estimated value of
- the estimated value can be updated iteratively as follows:
- Step 3 Based on estimated value and Establish a dimensionality reduction observation model of the received signal;
- the received signal can be further expressed as:
- the number of pre-detection traces I is much smaller than the number of units corresponding to the distance Doppler spectrum, therefore, the dimension of the vector to be reconstructed can be greatly reduced by the formula (22), and the formula (22) is the dimensionality reduction observation model of the present invention. .
- Step 4) reconstructing the target vector based on the dimensionality reduction observation model and the sparse recovery algorithm
- the sparse reconstruction method can be selected from Generalized Approximate Message Passing (GAMP), sparse Bayesian learning method, etc.
- GAMP Generalized Approximate Message Passing
- the present invention reconstructs ⁇ based on the GAMP algorithm and formula (22).
- I obeys the Bernoulli-Gaussian distribution of independent and identical distribution, and its marginal probability density function is:
- ⁇ is the Dirac function
- ⁇ is the sparsity
- ⁇ and ⁇ q represent the mean and variance of the Gaussian component, respectively.
- ⁇ , ⁇ and ⁇ q are all unknown and can be iteratively learned by the EM (Expectation-Maximum) algorithm.
- Step 5 Based on the reconstruction results, a generalized maximum likelihood ratio detector is designed to perform target detection, and output the detection results and parameters.
- the present invention is based on Perform object detection. For the i-th trace, judging whether it is the target can be reduced to the following hypothesis testing problem:
- H 0 and H 1 are the two hypotheses in the hypothesis testing problem.
- H 0 indicates that the i-th trace is not the target, and H 1 indicates that it is the target.
- the detection method is:
- EBD Estimate-before-detection
- Figure 2 shows the parameter estimation performance of the present invention before detection.
- the simulation parameters are shown in Table 1.
- Four targets are added to the received data, and three cases are considered: Case 1: The target is located in the integer Doppler unit, and the distance units are 100, 100.2, 140.3 and 191.5 respectively (the fractional part represents the size of the deviation from the integer unit.
- Case 2 The target is located in the integer distance unit, and the frequency units are 10, 15.2, 31.3 and 42.5; Case 3: The Doppler and distance of the target are deviated from the integer unit, respectively (15.2, 105.3), (31.2,, 140.4), (42.3, 191.1) and (53.5, 270.5); in the three cases, the estimation accuracy of the target parameters is measured by the mean square error (MSE), and the results are shown in Figure 2.
- MSE mean square error
- serial number parameter value 1 pulse interval 0.2ms 2 Pulse Width 25 ⁇ s 3 bandwidth 4MHz 4 Sampling Rate 5MHz 5 number of pulses 5 6 Linear processing of FFT points 64 7 Linear Processing Time Domain Window Function Hamming 8 Linear Processing Frequency Domain Window Function Chebyshev, -45dB
- Figure 3 presents the GAMP reconstruction results Statistical characteristics of the noise model value in the simulation, the parameters shown in Table 1 are still used in the simulation.
- Case 1 No target is included in the received signal
- Case 2 10 targets are added to the received signal, and the SNR is 0dB.
- Figure 3(a) and (b) give the QQplot curves of the noise samples in the two cases, respectively, and compare them with the standard Gaussian distribution.
- Figure 3(c) counts the correlation coefficients of the noise samples in the two cases. The results show that the modulo value of the output noise of the sparse reconstruction obeys an independent Gaussian distribution.
- the simulation experiment in FIG. 4 compares the performance of the EBD proposed by the present invention and the other three detectors, including the traditional detection method (TSPM, with windowing), the ideal matched filter (IMF, without windowing), and directly based on the formula (2)
- TSPM traditional detection method
- IMF ideal matched filter
- OGSR A method for direct sparse reconstruction and detection
- the parameters shown in Table 1 are still used, 10 targets are added in each simulation, and the target distance and Doppler are randomly generated.
- the simulation results are shown in Figure 4.
- the simulation results show that EBD has the ability of constant false alarm detection, and the detection performance gain is about 1.9dB compared with the conventional detection (TSPM) under the condition of 10 -5 false alarm rate.
- FIG. 5 and FIG. 6 verify the method of the present invention based on the measured data.
- the radar parameters corresponding to the data used are as follows: bandwidth 8MHz, sampling rate 10MHz, pulse width 12 microseconds, single frame pulse number 128, FFT point number 128, repetition frequency 12.5KHz.
- two DJI Phantom 3 UAVs were used as cooperation targets, and one frame was selected for analysis.
- the range Doppler spectrum is shown in Figure 5(a).
- the current frame beam points to no other target in the airspace. Since the SNRs of the two UAVs are relatively high, which are 26.3dB and 21.3dB respectively, in order to verify the performance of the proposed algorithm, noise is added to the original echo data.
- Object detection performance verification For example, in the original signal, after the noise power is increased by 6dB, the range Doppler spectrum with conventional processing is shown in Fig. 5(b).
- EBD has stronger anti-noise ability than TSPM, that is, has better object detection performance.
- EBD has a performance gain of 2.2dB and 1.5dB, respectively, under the condition of a false alarm rate of 10-5 .
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
一种基于检测前估计的雷达目标检测方法,包括以下步骤:基于常规脉冲多普勒处理与预检测,获取感兴趣点迹先验信息;对点迹距离和速度参数进行估计,在检测前完成参数估计;基于感兴趣点迹参数,建立接收信号的降维观测模型;基于稀疏恢复算法重构降维观测模型中的目标向量;基于重构结果,设计广义似然比检测器,进行目标检测,能够显著降低雷达信号处理损失,且目标检测器具有恒虚警特性,可大幅提升微弱目标检测性能。
Description
本发明涉及雷达信号处理领域,具体涉及一种基于检测前估计的雷达目标检测方法。
常规雷达信号处理通常基于匹配滤波器,采用先检测后估计的架构,为了降低空、时、频三个域强散射点副瓣对微弱目标检测的影响,需在滤波前对原始数据进行加窗。同时,由于目标参数未知,滤波器参数无法与真实目标相匹配。上述因素导致匹配滤波存在失配损失,大幅降低了雷达目标检测性能。
基于稀疏恢复的方法为解决上述问题提供了可行的思路。通过将目标参数空间离散化,构建完备观测矩阵A
od,接收回波信号可以表示为y=A
odx+n,x表示待重构目标向量,n表示噪声。那么,通过求解x即可实现对目标的重构及参数估计。通常,x是稀疏的,求解x可以归结为稀疏恢复问题。然而,基于稀疏重构方法进行目标检测尚面临三个方面的挑战:超高维度条件下的低复杂度重构、离网格下的精确重构、基于重构结果的恒虚警检测。
在超高维度条件下的低复杂度处理研究方面,通常,雷达目标探测所面临的信号模型维度巨大,低复杂度算法成为稀疏重构在雷达目标检测中应用的关键。迭代阈值类算法具有较低的复杂度,研究已经比较成熟。然而,此类方法收敛速度很慢,且迭代过程中,噪声(真实与估计值之间的差)逐渐偏离高斯分布,导致目标重构性能较差。另外,由于噪声的统计特性分析困难,难以基于重构结果进行CFAR(Constant False Alarm Rate,恒虚警)检测。近似消息传递(Approximate Message Passing,AMP)类方法在保持相对较低的复杂度同时,提升了目标重构精度,然而,应用于雷达目标检测与估计问题时,仍然需要降维。
离网格下的稀疏恢复方法研究方面,当前,解决离网格问题主要有两种方法。一种是通过在观测矩阵上引入观测误差,建立真实观测模型。模型中的观测矩阵与目标向量均未知,基于交替迭代的方法分别进行求解。显然,迭代过程大幅增加稀疏恢复算法复杂度。另一种是原子范数类方法,此类方法涉及超高维矩阵分解、矩阵求逆等,算法复杂度更加难以承受,在雷达目标检测与估计问题中难以使用。
基于重构结果的目标检测方面,实际上,目前的研究主要集中在重构精度方面,对 于虚警率和检测概率关注较少。虽然大多重构算法本身包括非线性去噪,但仍然没有直接给出判决信息,需要进一步基于重构结果进行目标检测。由于去噪函数的非线性,重构输出噪声的统计特性与探测场景密切相关,基于重构输出进行目标恒虚警检测面临众多困难。基于AMP方法,有关学者研究了目标检测问题,研究结果表明,在大系统极限条件下,AMP迭代过程中噪声仍然服从高斯分布,进而使得CFAR检测器设计成为可能。然而,如何在降维条件下,进行有效目标检测,尚缺乏相关方法。
发明内容
发明目的:针对常规基于稀疏重构的目标检测方法面临的挑战,本发明提供一种基于检测前估计的雷达目标检测方法。
技术方案:一种基于检测前估计的雷达目标检测方法,包括以下步骤:
4)基于稀疏恢复算法对降维观测模型中的目标向量进行重构;
5)基于重构结果,利用广义最大似然比检测器进行目标检测,输出目标检测结果和参数。
有益效果:本发明在基于点迹参数估计结果建立降维观测模型后,进一步利用目标稀疏先验信息,采用稀疏恢复算法进行目标重构和检测,解决了基于稀疏恢复的目标检测方法面临的三个关键问题,即超高维度条件下的降维、离网格条件下的精确重构和基于非线性重构结果的恒虚警目标检测。本发明在检测前完成目标参数估计,使得“匹配”信号处理成为可能,可大幅降低信号处理损失,提升目标检测性能。仿真和实测数据验证结果表明,与常规方法相比,本发明目标检测性能改善大于1.5dB。
图1为本发明的基于检测前估计的雷达目标检测方法流程图;
图2为本发明实施例提供的点迹参数估计精度示意图;
图3为本发明实施例提供的基于降维模型的重构输出噪声统计特性示意图;
图4为本发明实施例提供的仿真条件下4种检测器目标检测性能对比示意图;
图5为本发明实施例提供的实测数据距离多普勒谱示意图;
图6为实测数据条件下本发明方法与常规方法性能对比示意图。
下面结合附图对本发明的技术方案作进一步说明。
本发明适用于采用线性调频脉冲的脉冲体制雷达,目标检测处理流程如图1所示,包括以下步骤:
步骤1)对接收的基带数据,进行常规脉冲压缩和相参积累,得到距离多普勒谱,基于距离多普勒谱进行预检测,获得感兴趣点迹;
基带数据可以包括快时间、慢时间和空域数据,本发明适用于对三个域数据进行处理。因空域数据模型与慢时间域类似,本发明以快时间和慢时间域为例进行介绍。假设一个相参处理间隔中,距离单元数为L,脉冲数为K,接收信号可以表示为矩阵
忽略距离多普勒走动影响,其元素可以表示为:
其中y
l,k为第l个距离单元、第k个脉冲回波信号,0≤k≤K-1,0≤l≤L-1。P为目标数,T
s为采样时间间隔,α
p为第p个目标幅度,r
p为目标距离,T
pul为脉冲持续时间,T
I为脉冲间隔,μ为调频斜率,c为光速。f
d,p=2υ
p/λ为多普勒频移,υ
p为目标速度,λ为载波波长。n
l,k为加性高斯白噪声。j表示虚数,rect表示矩形函数。
在快时间域进行脉冲压缩、慢时间进行相参积累,可获得距离多普勒谱。然后,基于距离多普勒谱,采用常规自适应恒虚警检测方法,进行目标预检测,获得感兴趣点迹,所在距离单元和多普勒单元对应的距离和多普勒频率分别用
和
表示。
其中,y表示一个相参处理间隔的接收信号,由Y按列排列成一个向量得到,β为降维目标向量,其第i元素β
i表示第i个点迹的真实复幅度,如果第i个点迹为虚警,则β
i=0。n为加性高斯白噪声向量。
为近似的降维观测矩阵,其第i列为:
上标T表示转置。
其中,
通常,β仍然是稀疏的,可基于稀疏恢复对β进行重构,假设重构结果用
表示,那么,基于
即可实现目标检测。然而,实际应用中点迹并不位于整数单元,即
和
与点迹真实值存在偏差,直接基于式(2)进行重构面临离网格问题。对此,本发明首先基于
和
对点迹真实的参数
和
进行估计,然后再进行目标重构和检测。
其中,
符号;表示连接两个向量,形成一个向量。
的第i列为
是未知的,但显然
与
很接近,即真实目标参数与预检测点迹所在整数单元对应的参数相差不大。令
是已知的。既然
与
很接近,那么
和
也很接近,即
很小。因此,本发明中考虑基于
估计
进一步可得到观测矩阵
具体过程如下:
其中,θ=[r;f],r和f分别表示距离和多普勒频率;最小化式(8)目标函数,可得:
则,式(8)中的目标函数可进一步表示为:
式(13)中需要求g(θ)求的一阶导数向量和二阶导数矩阵,计算非常复杂,下面进一步给出一种简化求解方法。
由(7)可得:
ω
c(t)为该窗函数在时间T
pul内的连续形式。归一化的情况下,显然有,
实际上,式(14)是在快时间和慢时间域对数据进行加权,通过加权可以实现第i个点迹与回波中的其它点迹解耦合,即第i个点迹与其它点迹近似不存在相互影响。那么,g(θ)可进一步近似表示为:
其中,η
i=[r
i,f
i]
T,r
i、f
i分别表示第i个点迹的距离和多普勒频率,a
i表示第i个点迹的导向矢量,通过公式(3)计算,A
\i表示去掉矩阵A的第i列,β
\i表示去掉向量I的第i个元素。那么,最小化g(θ)必须使得
最小。通过最小化u
i(η
i)可以获得第i点迹所对应真实参数
的估计值
最小化u
i(η
i)可得β
i的估计为:
将式(18)带入u
i(η
i),那么,最小化u
i(η
i)等价于最小化下式:
通常,预检测点迹数目I远小于距离多普勒谱所对应的单元数,因此,式(22)可大幅降低待重构向量的维度,式(22)即为本发明的降维观测模型。
步骤4)基于降维观测模型和稀疏恢复算法,对目标向量进行重构;
稀疏重构方法可选择广义消息传递方法(Generalized Approximate Message Passing,GAMP)、稀疏贝叶斯学习方法等,本发明基于GAMP算法和式(22)重构β。此处,假设I服从独立同分布的伯努利-高斯分布,其边缘概率密度函数为:
其中,δ为狄拉克函数,ρ为稀疏度,κ和τ
q分别表示高斯分量的均值和方差。ρ、κ和τ
q均是未知的,可以通过EM(Expectation-Maximum)算法迭代学习得到。
步骤5)基于重构结果,设计广义最大似然比检测器,进行目标检测,输出检测结果和参数。
基于广义似然比准则,经推导和近似,对于第i个点迹,检测方法为:
上面描述了本发明所提的基于检测前估计的雷达目标检测方法的实施步骤,为了验证该方法的有效性,下面通过仿真和实测试验进行说明。下文中本发明的方法简称为EBD(Estimate-before-detection)。图2给出了本发明检测前的参数估计性能。仿真参数如表1所示,接收数据中加入4个目标,考虑三种情况:情况1:目标位于整数多普 勒单元,距离单元分别为100,100.2,140.3和191.5(小数部分表示偏离整数单元的大小);情况2:目标位于整数距离单元,频率单元分别为10,15.2,31.3和42.5;情况3:目标的多普勒和距离均偏离整数单元,分别为(15.2,105.3),(31.2,,140.4),(42.3,191.1)和(53.5,270.5);三种情况下目标参数估计精度均用均方误差(MSE)衡量,结果如图2所示。结果表明,本发明检测前参数估计方法可以有效估计出目标的距离和多普勒。
表1雷达仿真参数
| 序号 | 参数 | 取值 |
| 1 | 脉冲间隔 | 0.2ms |
| 2 | 脉冲宽度 | 25μs |
| 3 | 带宽 | 4MHz |
| 4 | 采样率 | 5MHz |
| 5 | 脉冲数 | 5 |
| 6 | 线性处理FFT点数 | 64 |
| 7 | 线性处理时域窗函数 | 汉明 |
| 8 | 线性处理频域窗函数 | 切比雪夫,-45dB |
图3给出了GAMP重构结果
中噪声模值的统计特性,仿真中仍然采用表1所示参数。考虑2种情况:情况1:接收信号中不包含目标;情况2:接收信号中加入10个目标,SNR均为0dB。统计噪声特性时,去除真实目标样本,即仅统计噪声样本。图3(a)和(b)分别给出了两种情况下噪声样本的QQplot曲线,并与标准高斯分布进行比较。图3(c)统计了两种情况下噪声样本的相关系数。结果表明,稀疏重构输出噪声的模值服从独立高斯分布。
图4的仿真实验对比了本发明所提的EBD与另外3种检测器的性能,包括传统检测方法(TSPM,加窗)、理想匹配滤波(IMF,不加窗)、直接基于式(2)直接进行稀疏重构和检测的方法(OGSR)。仿真中,仍然采用表1所示参数,每次仿真加入10个目标,目标距离和多普勒随机产生,仿真结果如图4所示。仿真结果表明,EBD均有恒虚警检测能力,在虚警率为10
-5条件下,与常规检测(TSPM)相比,检测性能增益约为1.9dB。
图5和图6基于实测数据对本发明方法进行了验证。所采用数据对应的雷达参数如下:带宽8MHz,采样率10MHz,脉冲宽度12微秒,单帧脉冲数128个,FFT点数128, 重复频率12.5KHz。试验中,采用两架大疆精灵3无人机作为合作目标,选择其中1帧进行分析,距离多普勒谱如图5(a)所示。当前帧波束指向空域无其它目标。由于两架无人机SNR较高,分别为26.3dB和21.3dB,为了验证提出算法的性能,在原始回波数据中加入噪声,在不同的噪声功率增加量情况下,基于蒙特卡洛仿真进行目标检测性能验证。例如,在原始信号中,噪声功率增加6dB后,采用常规处理的距离多普勒谱如图5(b)所示。
Claims (5)
- 根据权利要求1所述的基于检测前估计的雷达目标检测方法,其特征在于,所述步骤2)包括:其中,θ=[r;f],最小化式(8)目标函数,得:则,式(8)中的目标函数表示为:
- 利用窗函数将式(7)表示为:其中, w d表示慢时间域的归一化窗函数, 表示第i(i=1,2,...,I)个预检测点迹的快时间域归一化窗函数,I表示预检测点迹数目;归一化的情况下,显然有 那么,基于最小二乘准则, 和β的估计分别为:g(θ)进一步表示为:其中,η i=[r i,f i] T,r i、f i分别表示第i个点迹的距离和多普勒频率,a i表示第i个点迹的导向矢量,A \i表示去掉矩阵A的第i列,β \i表示去掉向量β的第i个元素,那么,最小化g(θ)必须使得 最小,通过最小化u i(η i)即能够获得第i个点迹所对应真实参数 的估计值最小化u i(η i),得到β i的估计为:将式(18)带入u i(η i),那么,最小化u i(η i)等价于最小化下式:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US17/732,531 US12235344B2 (en) | 2020-07-28 | 2022-04-29 | Radar target detection method based on estimation before detection |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010736841.4A CN111965615B (zh) | 2020-07-28 | 2020-07-28 | 一种基于检测前估计的雷达目标检测方法 |
| CN202010736841.4 | 2020-07-28 |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| US17/732,531 Continuation US12235344B2 (en) | 2020-07-28 | 2022-04-29 | Radar target detection method based on estimation before detection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2022022239A1 true WO2022022239A1 (zh) | 2022-02-03 |
Family
ID=73363262
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2021/104677 Ceased WO2022022239A1 (zh) | 2020-07-28 | 2021-07-06 | 一种基于检测前估计的雷达目标检测方法 |
Country Status (3)
| Country | Link |
|---|---|
| US (1) | US12235344B2 (zh) |
| CN (1) | CN111965615B (zh) |
| WO (1) | WO2022022239A1 (zh) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114563777A (zh) * | 2022-02-21 | 2022-05-31 | 中国科学院声学研究所 | 基于二维离网稀疏学习的时延多普勒频率联合估计方法 |
| CN114755654A (zh) * | 2022-06-14 | 2022-07-15 | 中达天昇(江苏)电子科技有限公司 | 一种基于图像拟态技术的残损雷达信号修复方法 |
| CN114994631A (zh) * | 2022-06-27 | 2022-09-02 | 中国民航大学 | 一种非凸松弛原子范数空时动目标参数估计方法 |
| CN117491986A (zh) * | 2023-10-11 | 2024-02-02 | 中通服咨询设计研究院有限公司 | 一种基于低重频ofdm雷达的远距高速目标跟踪方法 |
| CN119738781A (zh) * | 2024-11-15 | 2025-04-01 | 清华大学 | 一种基于深度展开的雷达恒虚警检测方法和系统 |
| CN121541167A (zh) * | 2026-01-16 | 2026-02-17 | 浙江大学 | 一种针对脉冲多普勒雷达的复杂电磁干扰环境下多目标检测方法及装置 |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112882005B (zh) * | 2021-01-14 | 2024-01-26 | 东南大学 | 一种基于贝叶斯学习的otfs雷达目标参数估计方法 |
| CN113253224B (zh) * | 2021-04-03 | 2023-03-14 | 西安电子科技大学 | 基于近似消息传递算法的被动分布式雷达目标检测方法 |
| CN113759316B (zh) * | 2021-08-27 | 2023-10-27 | 江苏云禾峰智能科技有限公司 | 一种基于fft的线性调频连续波雷达检测前参数估计方法 |
| CN114296039B (zh) * | 2021-12-01 | 2022-07-26 | 南京航空航天大学 | 一种基于稀疏重构的lfmcw雷达目标恒虚警检测方法及设备 |
| CN116299407B (zh) * | 2023-03-28 | 2026-03-17 | 西南石油大学 | 基于量化的mimo雷达通信系统目标速度估计和性能评估方法 |
| CN116933006B (zh) * | 2023-06-06 | 2025-08-19 | 西安电子科技大学 | 基于同源多频和非凸替代函数的外辐射雷达目标检测方法 |
| CN117351440B (zh) * | 2023-12-06 | 2024-02-20 | 浙江华是科技股份有限公司 | 基于开放式文本检测的半监督船舶检测方法及系统 |
| CN117591889B (zh) * | 2024-01-18 | 2024-04-26 | 北京中科飞鸿科技股份有限公司 | 一种基于神经网络的探测队形优化方法 |
| CN118330599B (zh) * | 2024-06-13 | 2024-08-13 | 清华大学 | 兼容脉冲多普勒与频率捷变波形的雷达信号处理装置 |
| CN119829897B (zh) * | 2024-12-04 | 2025-09-12 | 中国科学院声学研究所 | 一种能够抑制失配信号的自适应目标检测器的设计方法 |
| CN119511255B (zh) * | 2024-12-12 | 2025-10-31 | 上海船舶电子设备研究所(中国船舶集团有限公司第七二六研究所) | 基于幅度起伏特性的水下目标回波检测方法及系统 |
| CN119758285B (zh) * | 2024-12-27 | 2026-04-21 | 西安电子科技大学 | 一种非高斯噪声环境下的无源探测方法、装置及电子设备 |
| CN120742244B (zh) * | 2025-08-28 | 2025-12-12 | 西安昕晨电子科技有限公司 | 一种基于多功能雷达的信号采集处理方法及系统 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102279387A (zh) * | 2011-07-18 | 2011-12-14 | 西安电子科技大学 | Mimo雷达的目标到达角估计方法 |
| CN103969629A (zh) * | 2014-04-21 | 2014-08-06 | 河海大学 | 一种基于主瓣杂波配准的机载雷达杂波自适应抑制方法 |
| CN104914421A (zh) * | 2015-06-08 | 2015-09-16 | 中国民航大学 | 基于和差波束的低空风切变风速估计方法 |
| CN106324596A (zh) * | 2016-09-09 | 2017-01-11 | 深圳大学 | 基于压缩感知的随机脉冲多普勒雷达角度‑多普勒成像方法 |
| US10620296B1 (en) * | 2017-01-03 | 2020-04-14 | Toyon Research Corporation | System for positioning, navigation, and time (PNT) |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6141373A (en) * | 1996-11-15 | 2000-10-31 | Omnipoint Corporation | Preamble code structure and detection method and apparatus |
| CN103197301B (zh) * | 2013-03-19 | 2014-11-19 | 中国人民解放军海军航空工程学院 | 海面微动目标Radon-线性正则变换长时间相参积累检测方法 |
| FR3051918B1 (fr) * | 2016-05-31 | 2020-02-28 | Thales | Procede de detection adaptative d'une cible par un dispositif radar en presence d'interferences stationnaires, et radar et autodirecteur de missile mettant en oeuvre un tel procede |
| CN106093908A (zh) * | 2016-08-09 | 2016-11-09 | 西安电子科技大学 | 一种基于分块分段aic模型的雷达目标检测方法 |
| WO2018049595A1 (zh) * | 2016-09-14 | 2018-03-22 | 深圳大学 | 一种基于交替方向乘子法的稳健稀疏恢复stap方法及其系统 |
| CN108931766B (zh) * | 2018-04-28 | 2022-02-01 | 河海大学 | 一种基于稀疏重构的非均匀stap干扰目标滤除方法 |
| CN108845316A (zh) * | 2018-06-04 | 2018-11-20 | 中国卫星海上测控部 | 一种基于压缩感知理论的雷达稀疏探测方法 |
| CN110895331A (zh) * | 2019-08-21 | 2020-03-20 | 常熟理工学院 | 基于结构化观测矩阵的脉冲多普勒雷达目标稀疏探测方法 |
| CN110646786A (zh) * | 2019-08-21 | 2020-01-03 | 常熟理工学院 | 一种压缩感知雷达多目标速度-距离配对观测方法 |
| CN111044996A (zh) * | 2019-12-20 | 2020-04-21 | 南京航空航天大学 | 一种基于降维近似消息传递的lfmcw雷达目标检测方法 |
| US11927664B2 (en) * | 2021-02-25 | 2024-03-12 | Nxp B.V. | Radar-based detection using angle of arrival estimation based on sparse array processing |
-
2020
- 2020-07-28 CN CN202010736841.4A patent/CN111965615B/zh active Active
-
2021
- 2021-07-06 WO PCT/CN2021/104677 patent/WO2022022239A1/zh not_active Ceased
-
2022
- 2022-04-29 US US17/732,531 patent/US12235344B2/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102279387A (zh) * | 2011-07-18 | 2011-12-14 | 西安电子科技大学 | Mimo雷达的目标到达角估计方法 |
| CN103969629A (zh) * | 2014-04-21 | 2014-08-06 | 河海大学 | 一种基于主瓣杂波配准的机载雷达杂波自适应抑制方法 |
| CN104914421A (zh) * | 2015-06-08 | 2015-09-16 | 中国民航大学 | 基于和差波束的低空风切变风速估计方法 |
| CN106324596A (zh) * | 2016-09-09 | 2017-01-11 | 深圳大学 | 基于压缩感知的随机脉冲多普勒雷达角度‑多普勒成像方法 |
| US10620296B1 (en) * | 2017-01-03 | 2020-04-14 | Toyon Research Corporation | System for positioning, navigation, and time (PNT) |
Non-Patent Citations (1)
| Title |
|---|
| YOU JUN: "Key Techniques of Multi-FM-Broadcasting-Based Passive Radar", CHINESE DOCTORAL DISSERTATIONS FULL-TEXT DATABASE, no. 7, 1 April 2015 (2015-04-01), XP055896267, ISSN: 1674-022X * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114563777A (zh) * | 2022-02-21 | 2022-05-31 | 中国科学院声学研究所 | 基于二维离网稀疏学习的时延多普勒频率联合估计方法 |
| CN114755654A (zh) * | 2022-06-14 | 2022-07-15 | 中达天昇(江苏)电子科技有限公司 | 一种基于图像拟态技术的残损雷达信号修复方法 |
| CN114755654B (zh) * | 2022-06-14 | 2022-11-18 | 中达天昇(江苏)电子科技有限公司 | 一种基于图像拟态技术的残损雷达信号修复方法 |
| CN114994631A (zh) * | 2022-06-27 | 2022-09-02 | 中国民航大学 | 一种非凸松弛原子范数空时动目标参数估计方法 |
| CN117491986A (zh) * | 2023-10-11 | 2024-02-02 | 中通服咨询设计研究院有限公司 | 一种基于低重频ofdm雷达的远距高速目标跟踪方法 |
| CN119738781A (zh) * | 2024-11-15 | 2025-04-01 | 清华大学 | 一种基于深度展开的雷达恒虚警检测方法和系统 |
| CN121541167A (zh) * | 2026-01-16 | 2026-02-17 | 浙江大学 | 一种针对脉冲多普勒雷达的复杂电磁干扰环境下多目标检测方法及装置 |
Also Published As
| Publication number | Publication date |
|---|---|
| US12235344B2 (en) | 2025-02-25 |
| CN111965615B (zh) | 2022-05-13 |
| CN111965615A (zh) | 2020-11-20 |
| US20230059515A1 (en) | 2023-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111965615B (zh) | 一种基于检测前估计的雷达目标检测方法 | |
| CN111142105B (zh) | 复杂运动目标isar成像方法 | |
| CN105699950B (zh) | 基于自适应迭代前后向平滑共轭梯度的雷达杂波抑制方法 | |
| CN113534151B (zh) | 基于离网稀疏贝叶斯学习的双频段isar成像方法 | |
| CN107329115B (zh) | 基于gcrbf网络的lfm信号参数估计方法 | |
| CN113238193A (zh) | 一种多分量联合重构的sar回波宽带干扰抑制方法 | |
| CN105911527B (zh) | 基于efa与mwf的机载雷达空时自适应处理方法 | |
| CN110515052A (zh) | 一种基于时间反演的超宽带频域非等间隔采样目标检测方法 | |
| CN114047481A (zh) | 一种基于子空间正交性的稳健自适应波束形成方法 | |
| CN115421115B (zh) | 一种用于联合相位校正与isar成像的重赋权交替方向乘子法 | |
| CN107831473A (zh) | 基于高斯过程回归的距离‑瞬时多普勒图像序列降噪方法 | |
| CN114609607A (zh) | 低复杂度mimo雷达sr stap海杂波抑制方法 | |
| CN110471026A (zh) | 一种相位增强的米波雷达目标低仰角doa估计方法 | |
| CN112255595A (zh) | 一种基于仿机载测量下的海杂波数据预处理方法 | |
| CN109541567B (zh) | 基于深度学习的高速机动目标检测方法 | |
| CN108919189B (zh) | 一种频率和方位联合估计的阵列信号处理方法 | |
| CN113466864B (zh) | 快速联合免逆稀疏贝叶斯学习超分辨isar成像算法 | |
| CN112230200A (zh) | 一种基于激光雷达回波信号的改进型组合降噪方法 | |
| Nareddy et al. | Some intriguing observations on the learnt matrices in deep unfolded networks | |
| Smith et al. | Improved moving target detection in homogeneous and heterogeneous clutter with limited secondary data using unit circle roots constraints | |
| CN114910905B (zh) | 相似性约束下geo星机双基sar动目标智能成像方法 | |
| CN111044996A (zh) | 一种基于降维近似消息传递的lfmcw雷达目标检测方法 | |
| CN114296039B (zh) | 一种基于稀疏重构的lfmcw雷达目标恒虚警检测方法及设备 | |
| CN113759316B (zh) | 一种基于fft的线性调频连续波雷达检测前参数估计方法 | |
| CN114879157B (zh) | 基于能量轨迹提取的高价值目标散射中心参数估计方法 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21850181 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 21850181 Country of ref document: EP Kind code of ref document: A1 |



