CN103399307B

CN103399307B - A kind of method of correction channel error and device

Info

Publication number: CN103399307B
Application number: CN201310337409.8A
Authority: CN
Inventors: 张磊; 王宇; 邓云凯
Original assignee: Institute of Electronics of CAS
Current assignee: Institute of Electronics of CAS
Priority date: 2013-08-05
Filing date: 2013-08-05
Publication date: 2015-10-28
Anticipated expiration: 2033-08-05
Also published as: CN103399307A

Abstract

The invention discloses a kind of method of correction channel error, described method comprises: first carry out amplitude correction and alignment to the echo data of each passage in the hyperchannel obtained; Then the described echo data after amplitude correction and alignment is carried out imaging processing and synthetic image, according to described image determination cost function, according to described cost function estimating channel phase error, and according to described channel phase errors, described echo data is corrected.The present invention also discloses a kind of device of correction channel error.Adopt technical scheme of the present invention, achieve complete, exactly estimate and correction channel error.

Description

Method and device for correcting channel error

技术领域technical field

本发明涉及通道误差估计和校正技术，尤其涉及一种校正通道误差的方法及装置。The invention relates to channel error estimation and correction technology, in particular to a method and device for correcting channel errors.

背景技术Background technique

多通道合成孔径雷达（SAR）由于系统设计、器件精度、工作环境等影响，通道的传输特性不可避免的存在不一致，这些不一致反映到回波数据上体现为通道误差。通道误差会对成像等数据应用产生极大影响，应给予校正。In multi-channel synthetic aperture radar (SAR), due to the influence of system design, device accuracy, and working environment, there are inevitably inconsistencies in the transmission characteristics of the channels, and these inconsistencies are reflected in the echo data as channel errors. Channel errors will have a great impact on data applications such as imaging and should be corrected.

目前，现有技术中关于通道误差估计和校正方法包括两类，一类是基于内定标的，另一类是基于回波数据的；其中，基于内定标的方法通过雷达的定标子系统获取定标信息并依此对通道误差进行校正，这种方法比较准确但无法监控定标回路外的部分如天线；基于回波数据的方法一般是将信号投影到子空间，这样可以大大降低噪声的影响；但是这类方法需要准确地知道系统参数，如通道的等效相位中心距离等，而这些系统参数一般是利用回波数据进行估计得到，系统参数估计的准确性会直接影响通道误差估计方法的准确性。At present, there are two types of channel error estimation and correction methods in the prior art, one is based on internal calibration, and the other is based on echo data; among them, the method based on internal calibration obtains the calibration through the calibration subsystem of the radar This method is more accurate but cannot monitor the part outside the calibration loop such as the antenna; the method based on echo data generally projects the signal to the subspace, which can greatly reduce the influence of noise; However, this kind of method needs to know the system parameters accurately, such as the channel’s equivalent phase center distance, etc., and these system parameters are generally estimated by using echo data, and the accuracy of system parameter estimation will directly affect the accuracy of the channel error estimation method sex.

综上所述，如何完整、准确地估计和校正通道误差是目前亟待解决的问题。To sum up, how to completely and accurately estimate and correct the channel error is an urgent problem to be solved.

发明内容Contents of the invention

有鉴于此，本发明的主要目的在于提供一种校正通道误差的方法及装置，能够完整、准确地估计和校正通道误差。In view of this, the main purpose of the present invention is to provide a method and device for correcting channel errors, which can completely and accurately estimate and correct channel errors.

为达到上述目的，本发明的技术方案是这样实现的：In order to achieve the above object, technical solution of the present invention is achieved in that way:

本发明提供了一种校正通道误差的方法，所述方法包括：对获取的多通道中每一通道的回波数据进行幅度校正和对齐；将幅度校正和对齐后的所述回波数据进行成像处理而生成图像，根据所述图像确定代价函数，根据所述代价函数估计通道相位误差，并根据所述通道相位误差对所述回波数据进行校正。The present invention provides a method for correcting channel errors, the method comprising: performing amplitude correction and alignment on the acquired echo data of each channel in multiple channels; imaging the echo data after amplitude correction and alignment processing to generate an image, determining a cost function based on the image, estimating a channel phase error based on the cost function, and correcting the echo data based on the channel phase error.

上述方案中，所述对获取的多通道中每一通道的回波数据进行幅度校正，包括：对每一所述通道的所述回波数据进行方位向幅度求平均，根据每一所述通道与参考通道的幅度平均值之比确定通道幅度误差；根据所述通道幅度误差对每一所述通道的所述回波数据进行补偿。In the above solution, the amplitude correction of the acquired echo data of each channel in the multi-channel includes: averaging the azimuth and amplitude of the echo data of each of the channels, and according to each of the channels A channel amplitude error is determined as a ratio to an amplitude average value of a reference channel; and the echo data of each of the channels is compensated according to the channel amplitude error.

上述方案中，所述对获取的多通道中每一通道的回波数据进行对齐，包括：分别对幅度校正后的每一通道的回波数据和参考通道的回波数据进行距离向快速傅里叶变换，然后将两者共轭相乘，并对共轭相乘后的所述回波数据在方位向求平均，确定求平均后的所述回波数据的相位，根据所述相位的斜率确定通道间的采样起始误差；将所述幅度校正后的每一通道的回波数据与相应的所述相位作积。In the above solution, the alignment of the acquired echo data of each channel in the multi-channel includes: respectively performing range-wise fast Fourier analysis on the echo data of each channel after amplitude correction and the echo data of the reference channel. leaf transformation, and then multiply the two conjugates, and average the echo data in the azimuth direction after the conjugate multiplication, determine the phase of the averaged echo data, according to the slope of the phase Determining a sampling start error between channels; multiplying the amplitude-corrected echo data of each channel with the corresponding phase.

上述方案中，所述方法包括：所述幅度校正和对齐交替迭代进行，确定估计的采样起始误差和/或通道幅度误差的变化值满足预设的判决门限时迭代结束。In the above solution, the method includes: the amplitude correction and alignment are performed alternately and iteratively, and the iteration ends when it is determined that the change value of the estimated sampling start error and/or channel amplitude error satisfies a preset decision threshold.

上述方案中，根据所述图像确定代价函数，包括：将所述图像在方位向进行自相关，并在距离向求平均，将求平均后的结果作为代价函数。In the above solution, determining the cost function according to the image includes: performing autocorrelation on the image in the azimuth direction, and averaging in the distance direction, and using the averaged result as the cost function.

上述方案中，对获取的多通道中每一通道的回波数据进行对齐之前，所述方法还包括：利用合成孔径雷达系统的条带成像模式获取多通道中每一通道的回波数据。In the above solution, before aligning the acquired echo data of each of the multi-channels, the method further includes: acquiring the echo data of each of the multi-channels by using a strip imaging mode of the synthetic aperture radar system.

本发明还提供了一种校正通道误差的装置，所述装置包括第一校正单元、成像单元、确定单元、估计单元和第二校正单元；其中，所述第一校正单元，用于对获取的多通道中每一通道的回波数据进行幅度校正和对齐；所述成像单元，用于将幅度校正和对齐后的所述回波数据进行成像处理而生成图像；所述确定单元，用于根据所述图像确定代价函数；所述估计单元，用于根据代价函数估计通道相位误差；所述第二校正单元，用于根据所述通道相位误差对所述回波数据进行校正。The present invention also provides a device for correcting channel errors, which includes a first correction unit, an imaging unit, a determination unit, an estimation unit, and a second correction unit; wherein, the first correction unit is used to correct the acquired The echo data of each channel in the multi-channel is subjected to amplitude correction and alignment; the imaging unit is used to perform imaging processing on the echo data after amplitude correction and alignment to generate an image; the determination unit is used to generate an image according to The image determines a cost function; the estimation unit is configured to estimate a channel phase error according to the cost function; the second correction unit is configured to correct the echo data according to the channel phase error.

上述方案中，所述第一校正单元进一步用于：对每一所述通道的所述回波数据进行方位向幅度求平均，根据每一所述通道与参考通道的幅度的平均值之比确定通道幅度误差；根据所述通道幅度误差对每一所述通道的所述回波数据进行补偿。In the above scheme, the first correction unit is further configured to: average the azimuth and amplitude of the echo data of each of the channels, and determine according to the ratio of the average value of the amplitude of each of the channels to the reference channel Channel amplitude error: Compensating the echo data of each channel according to the channel amplitude error.

上述方案中，所述第一校正单元进一步用于：对幅度校正后的每一通道的回波数据和参考信道的回拨数据进行距离向快速傅里叶变换，然后将两者共轭相乘，并对共轭相乘后的所述回波数据在方位向求平均，确定求平均后的所述回波数据的相位，根据所述相位的斜率确定通道间的采样起始误差；将所述幅度校正后的每一通道的回波数据与相应的所述相位作积。In the above scheme, the first correction unit is further configured to: perform range-to-fast Fourier transform on the amplitude-corrected echo data of each channel and the callback data of the reference channel, and then conjugate multiply the two , and average the echo data after conjugate multiplication in the azimuth direction, determine the phase of the echo data after averaging, and determine the sampling start error between channels according to the slope of the phase; The amplitude-corrected echo data of each channel is multiplied with the corresponding phase.

上述方案中，所述确定单元进一步用于：将所述图像在方位向进行自相关，并在距离向求平均，将求平均后的结果作为代价函数。In the above solution, the determining unit is further configured to: perform autocorrelation on the image in the azimuth direction, and calculate an average in the distance direction, and use the averaged result as a cost function.

本发明提供的校正通道误差的方法及装置，先对获取的多通道中每一通道的回波数据进行幅度校正和对齐；然后将幅度校正和对齐后的所述回波数据进行成像处理而生成图像，根据所述图像确定代价函数，根据所述代价函数估计通道相位误差，并根据所述通道相位误差对所述回波数据进行校正；如此，本发明基于成像质量确定代价函数，然后根据代价函数估计通道相位误差，从而本发明无需准确的知道系统参数就能够对通道误差的估计和校正。The method and device for correcting channel errors provided by the present invention first perform amplitude correction and alignment on the acquired echo data of each channel in multiple channels; then perform imaging processing on the echo data after amplitude correction and alignment to generate images, determine the cost function according to the image, estimate the channel phase error according to the cost function, and correct the echo data according to the channel phase error; thus, the present invention determines the cost function based on the imaging quality, and then according to the cost The function estimates the channel phase error, so that the present invention can estimate and correct the channel error without accurately knowing the system parameters.

附图说明Description of drawings

图1为本发明校正通道误差的方法的实现流程示意图；Fig. 1 is a schematic diagram of the implementation flow of the method for correcting channel errors of the present invention;

图2本发明实施例中估计通道幅度误差的平均幅度示意图；Fig. 2 is a schematic diagram of the average magnitude of the estimated channel magnitude error in an embodiment of the present invention;

图3本发明实施例中估计采样起始误差的线性相位示意图；Fig. 3 is a schematic diagram of a linear phase of an estimated sampling start error in an embodiment of the present invention;

图4为本发明实施例中估计通道相位误差的代价函数示意图；4 is a schematic diagram of a cost function for estimating a channel phase error in an embodiment of the present invention;

图5为本发明实施例中通道误差校正前成像结果示意图；5 is a schematic diagram of imaging results before channel error correction in an embodiment of the present invention;

图6为本发明实施例中通道误差校正后成像结果示意图；6 is a schematic diagram of imaging results after channel error correction in an embodiment of the present invention;

图7为本发明校正通道误差的装置的组成结构示意图。FIG. 7 is a schematic diagram of the composition and structure of the device for correcting channel errors of the present invention.

具体实施方式Detailed ways

根据误差的来源一般将通道误差分解为：通道幅度误差、采样起始误差和通道相位误差三部分。其中，通道幅度误差是由于各通道在距离向上天线方向图的不一致而造成的；采样起始误差是由于各通道采样时钟的不同步而造成的；各通道传输特性的不同造成了通道相位误差。这些误差叠加在通道m的回波信号上，该回波信号的表达式为：According to the source of the error, the channel error is generally decomposed into three parts: channel amplitude error, sampling start error and channel phase error. Among them, the channel amplitude error is caused by the inconsistency of the antenna pattern of each channel in the distance; the sampling start error is caused by the asynchronous sampling clock of each channel; the channel phase error is caused by the different transmission characteristics of each channel. These errors are superimposed on the echo signal of channel m, the expression of which echo signal is:

${s the s}_{m m} ((τ τ,, η η)) = = Δ Δ {a a}_{m m} ((τ τ)) \cdot \cdot {s the s}_{11} ((τ τ + + Δ Δ {τ τ}_{m m},, η η + + \frac{Δ Δ {D D.}_{m m}}{{v v}_{p p}})) \cdot \cdot exp exp ((jΔ jΔ {φ φ}_{m m})) - - - - - - ((11));;$

公式（1）中，τ为快时间即距离向上的时间，η为慢时间即方位向上的时间，Δa_m(τ)为通道m通道幅度误差，Δτ_m为通道m采样起始误差，ΔD_m为通道m的等效相位中心与参考通道的等效相位中心之间的距离，v_p为雷达平台的运动速度，Δφ_m为通道m通道相位误差，j为虚数，exp为以自然对数e为底的指数函数；s_m(τ,η)为以τ和η为自变量的通道m的SAR回波信号，下标m为合成孔径雷达的通道m；为以τ和η为自变量的通道1的SAR回波信号，Δτ_m和分别表示τ和η的时延，下标1为合成孔径雷达的通道1，其中SAR回波信号的表达式可参见公式（2）来确定。In the formula (1), τ is the fast time, that is, the time when the distance is upward, η is the slow time, that is, the time when the azimuth is upward, Δa _m (τ) is the channel amplitude error of channel m, Δτ _m is the sampling start error of channel m, ΔD _m is the distance between the equivalent phase center of the channel m and the equivalent phase center of the reference channel, v _p is the moving speed of the radar platform, Δφ _m is the phase error of the channel m, j is an imaginary number, exp is the natural logarithm e is an exponential function with the base; s _m (τ, η) is the SAR echo signal of the channel m with τ and η as independent variables, and the subscript m is the channel m of the synthetic aperture radar; is the SAR echo signal of channel 1 with τ and η as independent variables, Δτ _m and Denote the time delays of τ and η, respectively, and the subscript 1 is the channel 1 of the synthetic aperture radar, where the expression of the SAR echo signal can be determined by referring to formula (2).

下面以s₁(τ,η)为例，说明SAR回波信号的表达式，多通道合成孔径雷达参考通道（一般为合成孔径雷达的通道1）接收到的回波信号为：The following takes s ₁ (τ, η) as an example to illustrate the expression of the SAR echo signal. The echo signal received by the reference channel of the multi-channel synthetic aperture radar (generally channel 1 of the synthetic aperture radar) is:

s₁(τ,η)=Aw_r(τ-2R(η)/c)w_a(η-η_c)s ₁ (τ,η)=Aw _r (τ-2R(η)/c)w _a (η-η _c )

×exp(-j4πf₀R(η)/c)exp{jπK_r(τ-2R(η)/c)²} （2）；×exp(-j4πf ₀ R(η)/c)exp{jπK _r (τ-2R(η)/c) ² } (2);

公式（2）中，τ为快时间即距离向上的时间，η为慢时间即方位向上的时间，A为一个复常量，w_r为距离向窗函数，R(η)为瞬时斜距，c为光速，w_a为方位向窗函数，η_c为波束中心偏离时间，w_r为距离向窗函数，w_a为方位向窗函数，f₀为雷达载频，K_r为调频率，R(η)为瞬时斜距。In the formula (2), τ is the fast time, that is, the time when the distance is upward, η is the slow time, that is, the time when the azimuth is upward, A is a complex constant, w _r is the range window function, R(η) is the instantaneous slope distance, c is the speed of light, w _a is the azimuth window function, η _c is the beam center deviation time, w _r is the range window function, w _a is the azimuth window function, f ₀ is the radar carrier frequency, K _r is the modulation frequency, R( η) is the instantaneous slope distance.

本发明中假设合成孔径雷达的通道数量一共有N个，其中1≤m≤N，且指定通道1为参考通道。In the present invention, it is assumed that the SAR has N channels in total, where 1≤m≤N, and channel 1 is designated as the reference channel.

图1为本发明校正通道误差的方法的实现流程示意图，如图1所示，本发明校正通道误差的方法包括：Fig. 1 is a schematic diagram of the implementation process of the method for correcting channel errors of the present invention. As shown in Fig. 1, the method for correcting channel errors of the present invention includes:

步骤101：对获取的多通道中每一通道的回波数据进行幅度校正和对齐；Step 101: performing amplitude correction and alignment on the acquired echo data of each channel in the multi-channel;

这里，所述对获取的多通道中每一通道的回波数据进行幅度校正，包括以下步骤：针对通道m的通道误差估计，首先对回波数据的幅度在方位向上取平均，然后确定每一所述通道与参考通道的幅度平均值之比，所述幅度平均值之比为通道m的通道幅度误差Δa_m(τ)：Here, the amplitude correction of the acquired echo data of each channel in the multi-channel includes the following steps: for the channel error estimation of the channel m, first average the amplitude of the echo data in the azimuth direction, and then determine each The ratio of the average value of the amplitude of the channel to the reference channel, the ratio of the average value of the amplitude is the channel amplitude error Δam (τ) of the channel _m :

$Δ Δ {a a}_{m m} ((τ τ)) = = \frac{{E E.}_{η η} {{| | {s the s}_{m m} ((τ τ,, η η)) | |}}}{{E E.}_{η η} {{| | {s the s}_{11} ((τ τ + + Δ Δ {τ τ}_{m m},, η η + + \frac{Δ Δ {D D.}_{m m}}{{v v}_{p p}})) \cdot \cdot exp exp ((jΔ jΔ {φ φ}_{m m})) | |}}}$

$= = \frac{{E E.}_{η η} {{| | {s the s}_{m m} ((τ τ,, η η)) | |}}}{{E E.}_{η η} {{| | {s the s}_{11} ((τ τ + + Δ Δ {τ τ}_{m m},, η η)) | |}}}$

$= = \frac{{E E.}_{η η} {{| | {s the s}_{m m} ((τ τ,, η η)) | |}}}{{E E.}_{η η} {{| | {s the s}_{11} ((τ τ,, η η)) | |}}} - - - - - - ((33));;$

公式（3）中，E_η为按慢时间求期望，||为求幅度的符号，由于此时还不能确定Δτ_m，因此，在估计时分母用s₁(τ,η)替代s₁(τ+Δτ_m,η)。In the formula (3), E _η is to find the expectation according to the slow time, and || is the symbol to find the amplitude. Since Δτ _m cannot be determined at this time, the denominator is replaced by s ₁ (τ,η) in the estimation of s ₁ ( τ+Δτ _m , η).

这里，所述对获取的多通道中每一通道的回波数据进行对齐，包括以下步骤：对通道幅度误差校正后的通道m的回波数据和参考通道的回波数据进行距离向快速傅里叶变换（FFT，Fast Fourier Transformation），然后将两者共轭相乘，接着在方位向取平均，最后对结果取相位，可得公式（4）：Here, the alignment of the acquired echo data of each channel in the multi-channel includes the following steps: performing range-wise fast Fourier on the echo data of the channel m after channel amplitude error correction and the echo data of the reference channel leaf transformation (FFT, Fast Fourier Transformation), and then multiply the two conjugates, then take the average in the azimuth direction, and finally take the phase of the result, the formula (4) can be obtained:

$arg arg (({E E.}_{η η} {{{S S}_{11}^{* *} (({f f}_{r r},, η η)) \cdot \cdot {S S}_{m m} (({f f}_{r r},, η η))}})) = = 22 π π {f f}_{dc dc} \frac{Δ Δ {D D.}_{m m}}{{v v}_{p p}} + + 22 π π {f f}_{r r} Δ Δ {τ τ}_{m m} + + Δ Δ {φ φ}_{m m} - - - - - - ((44));;$

公式（3）中，arg为取复数的辐角，*为取共轭，S₁(f_r,η)为s₁(τ,η)变换到距离频域f_r的信号，S_m(f_r,η)为s_m(τ,η)变换到距离频域f_r的信号，f_r为距离频率，f_dc为多普勒中心，2πf_rΔτ_m为线性相位，2πf_rΔτ_m的斜率对应于采样起始误差，因此，对所述线性相位进行校正即可完成对回波数据的对齐。然后，再将对齐后的回波数据进行距离向逆快速傅里叶变换（IFFT），重新回到距离时间域。In formula (3), arg is the argument of the complex number, * is the conjugate, S ₁ (f _r ,η) is the signal transformed from s ₁ (τ,η) to the distance frequency domain f _r , S _m (f _r , η) is the signal transformed from s _m (τ, η) to the range frequency domain f _r , f _r is the range frequency, f _dc is the Doppler center, 2πf _r Δτ _m is the linear phase, and the slope of 2πf _r Δτ _m Corresponding to the sampling start error, therefore, the alignment of the echo data can be completed by correcting the linear phase. Then, the aligned echo data is subjected to a range-inverse fast Fourier transform (IFFT) to return to the range-time domain.

上述对齐操作将回波数据进行了包络平移，这会使通道间再次出现通道幅度误差，而且公式（3）中在估计Δa_m(τ)时使用s₁(τ,η)替代了分母s₁(τ+Δτ_m,η)；因此，幅度误差校正和采样起始误差校正应当交替迭代进行，即：估计通道幅度误差—校正通道幅度误差—估计采样起始误差—校正采样起始误差—估计通道幅度误差—校正通道幅度误差—估计采样起始误差—校正采样起始误差……当估计出的采样起始误差和/或通道幅度误差的变化值满足预设的门限要求时迭代结束。The above alignment operation performs envelope translation on the echo data, which will cause channel amplitude errors between channels again, and in formula (3), s ₁ (τ,η) is used to replace the denominator s when estimating Δa _m (τ) ₁ (τ+Δτ _m ,η); therefore, amplitude error correction and sampling start error correction should be performed alternately and iteratively, namely: estimate channel amplitude error—correct channel amplitude error—estimate sampling start error—correct sampling start error— Estimated channel amplitude error—corrected channel amplitude error—estimated sampling start error—corrected sampling start error...The iteration ends when the estimated sampling start error and/or the change value of the channel amplitude error meets the preset threshold requirement.

从上述的交替迭代可以看出：所述对获取的多通道中每一通道的回波数据进行幅度校正和对齐，可以先对回波数据进行幅度校正，然后再对幅度校正后的回波数据进行对齐；也可以先对回波数据进行对齐，然后再对对齐后的回波数据进行幅度校正；因此，在本发明幅度校正和对齐的先后顺序并不影响最后的通道误差的估计和校正的结果。From the above alternate iterations, it can be seen that the amplitude correction and alignment of the acquired echo data of each channel in the multi-channel can be performed on the echo data first, and then the amplitude corrected echo data It is also possible to align the echo data first, and then perform amplitude correction on the aligned echo data; therefore, the sequence of amplitude correction and alignment in the present invention does not affect the estimation and correction of the final channel error result.

步骤102：将幅度校正和对齐后的所述回波数据进行成像处理而生成图像，根据所述图像确定代价函数，根据所述代价函数估计通道相位误差，并根据所述通道相位误差对所述回波数据进行校正。Step 102: Perform imaging processing on the amplitude-corrected and aligned echo data to generate an image, determine a cost function according to the image, estimate a channel phase error according to the cost function, and calculate the channel phase error according to the channel phase error Echo data are corrected.

这里，所述根据所述图像确定代价函数，包括：将所述图像在方位向进行自相关，并在距离向求平均，将求平均后的结果作为代价函数。Here, the determining the cost function according to the image includes: autocorrelating the image in the azimuth direction and averaging in the distance direction, and using the averaged result as the cost function.

具体的，在雷达成像的自聚焦方法中，代价函数反映了雷达图像的成像质量，当成像质量最优时代价函数达到极值；在通道误差的估计和校正过程中，通道相位误差会造成图像的散焦，通过迭代可改变通道相位误差的估计值进而使得代价函数达到极值，因此，可以根据代价函数估计通道相位误差。Specifically, in the self-focusing method of radar imaging, the cost function reflects the imaging quality of the radar image, and the cost function reaches the extreme value when the imaging quality is optimal; in the process of channel error estimation and correction, the channel phase error will cause image The defocus of the channel phase error can be changed through iteration to make the cost function reach the extreme value. Therefore, the channel phase error can be estimated according to the cost function.

为了更好地体现通道相位误差对成像的影响，这里定义一种新的代价函数F_ISC(k)为：In order to better reflect the influence of channel phase error on imaging, a new cost function F _ISC (k) is defined here as:

${F f}_{ISC ISC} ((k k)) = = \frac{{E E.}_{p p} {{I I ((p p,, q q)) \cdot &Center Dot; {I I}^{* *} ((p p,, q q + + k k))}}}{{E E.}_{p p} {{I I ((p p,, q q)) \cdot \cdot {I I}^{* *} ((p p,, q q))}}} - - - - - - ((55));;$

该代价函数可称为沿图像方位向的自相关（ISC，Image Self Correlation），公式（5）中分子为自相关的平均，分母的作用为归一化；对于实际的雷达图像，当k=0时F_ISC(k)取得最大值1。具体的，I(p,q)为雷达图像，p和q分别为距离向坐标和方位向坐标，k为坐标偏移量，p、q和k均取整数，*为求共轭，E_p表示沿距离向求期望。This cost function can be called the autocorrelation along the image direction (ISC, Image Self Correlation). The numerator in the formula (5) is the average of the autocorrelation, and the denominator is the normalization; for the actual radar image, when k= When 0, F _ISC (k) takes the maximum value of 1. Specifically, I(p,q) is the radar image, p and q are the range coordinates and azimuth coordinates respectively, k is the coordinate offset, p, q and k are all integers, * is the conjugate, E _p Indicates seeking expectation along distance.

由于通道相位误差会使图像在方位向上产生虚假目标，这些虚假目标的与真实的目标相比，虽然存在一定的散焦，但其形状轮廓与真实目标是十分接近的。这样，根据数据自相关的特性，公式（5）定义的代价函数会在零点外存在数个峰值，通过这些峰值的抑制程度就可以准确地估计出通道相位误差。Due to the channel phase error, the image will produce false targets in the azimuth direction. Compared with the real target, these false targets have certain defocus, but their shape and outline are very close to the real target. In this way, according to the characteristics of data autocorrelation, the cost function defined by formula (5) will have several peaks outside the zero point, and the channel phase error can be accurately estimated by the degree of suppression of these peaks.

这里，所述根据代价函数估计通道相位误差也是通过迭代完成的，具体如下：Here, the estimation of the channel phase error according to the cost function is also done iteratively, as follows:

对于通道m先进行第一次估计，首先，从0～2π中任取一个相位值φ_m,1,1校正到通道m的回波数据上，将该回波数据成像并确定第一代价函数；然后，在φ_m,1,1的基础上再加上一个预设的相位值变为φ_m,1,2，再次将回波数据成像并确定第二代价函数，对比所述第一代价函数和所述第二代价函数形成的曲线：For channel m, the first estimation is performed first. First, a phase value φ _{m, 1, 1} is randomly selected from 0 to 2π and corrected to the echo data of channel m, and the echo data is imaged and the first cost function is determined. ; Then, add a preset phase value on the basis of φ _m,1,1 to change to φ _m,1,2 , image the echo data again and determine the second cost function, compared with the first cost function and the curve formed by the second cost function:

若第二代价函数形成的曲线中零点外的峰值被抑制，则继续增加固定的相位值，直到零点峰值有变大的趋势，此时校正的相位记为φ_m,1,n，而通道相位误差的真实值在φ_m,1,n-1至φ_m,1,n之间；If the peak outside the zero point in the curve formed by the second cost function is suppressed, continue to increase the fixed phase value until the peak value of the zero point tends to become larger. At this time, the corrected phase is recorded as φ _m,1,n , and the channel phase The true value of the error is between φ _m,1,n-1 and φ _m,1,n ;

若第二代价函数形成的曲线中零点峰值有变大趋势，则说明应该在φ_m,1,1基础上减去一定的相位，类似的可以得到区间φ_m,1,n-1至φ_m,1,n之间。If the zero-point peak in the curve formed by the second cost function tends to become larger, it means that a certain phase should be subtracted from φ _m , 1, 1. Similarly, the interval φ _{m, 1, n-1} to φ _m can be obtained _{, between 1 and n} .

在区间φ_m,1,n-1至φ_m,1,n内继续重复上述操作，这里，加减的预设相位值要随着区间的缩小而不断减小，根据要求的估计精度就可最终确定通道m的第一次估计的通道相位误差Δφ_m,1。Continue to repeat the above operations within the interval φ _m,1,n-1 to φ _m,1,n . Here, the preset phase value for addition and subtraction should decrease continuously as the interval shrinks. According to the required estimation accuracy, it can be The first estimated channel phase error Δφ _m,1 of channel m is finally determined.

根据通道m的第一次估计通道相位误差Δφ_m,1的过程，可以依次得到其余各通道的第一次的通道相位误差：Δφ_2,1、Δφ_3,1、Δφ_4,1……，根据这些估计值对回波数据进行校正后，再次从通道2开始进行第二次估计，依次得到第二估计的各通道的通道相位误差Δφ_2,2、Δφ_3,2、Δφ_4,2……；如此重复迭代，当所有估计得到的通道相位误差都小于需要的门限值时，迭代停止。According to the process of channel m's first estimated channel phase error Δφ _m,1 , the first channel phase errors of other channels can be obtained in turn: Δφ _2,1 , Δφ _3,1 , Δφ _4,1 ..., After the echo data are corrected according to these estimated values, the second estimation is performed again starting from channel 2, and the second estimated channel phase errors of each channel Δφ _2,2 , Δφ _3,2 , Δφ _4,2 ... . . . Repeat the iteration in this way, and when all the estimated channel phase errors are smaller than the required threshold value, the iteration stops.

前述根据代价函数估计通道相位误差的迭代过程中，相位φ_m,f,g中下标的含义为：第一个下标m表示通道m，第二个下标f表示第f次估计，第三个下标g表示所述第f次估计过程中取的第g个相位；具体的，相位φ_m,1,1中的第一个下标m表示通道m，第二个下标1表示第1次估计，第三个下标1表示所述第1次估计过程中取的第1个相位。相位φ_m,f中下标的含义为：第一个下标m表示通道m，第二个下标f表示第f次估计；具体的，相位φ_3,2中，第一个下标3表示通道3，第二个下标2表示第2次估计的相位。In the aforementioned iterative process of estimating the channel phase error based on the cost function, the meanings of the subscripts in the phase φ _{m, f, g} are as follows: the first subscript m represents the channel m, the second subscript f represents the fth estimation, and the third subscript The first subscript g represents the g-th phase taken in the f-th estimation process; specifically, the first subscript m in the phase φ _m,1,1 represents the channel m, and the second subscript 1 represents the first One estimation, the third subscript 1 indicates the first phase taken in the first estimation process. The meaning of subscripts in phase φ _m,f is: the first subscript m indicates channel m, and the second subscript f indicates the f-th estimation; specifically, in phase φ _3,2 , the first subscript 3 indicates Channel 3, the second subscript 2 indicates the second estimated phase.

在步骤101之前，所述方法还包括：采用合成孔径雷达系统的条带成像模式获取多通道中每一通道的回波数据。Before step 101, the method further includes: acquiring echo data of each channel in the multi-channels using a strip imaging mode of a synthetic aperture radar system.

以下结合实施例，对本发明校正通道误差的方法的具体实施作进一步详细阐述。The specific implementation of the method for correcting channel errors of the present invention will be further described in detail below in conjunction with the embodiments.

在本实施例中，合成孔径雷达系统有两个通道（通道1和通道2）工作在条带模式，其中以通道1为参考信道，合成孔径雷达系统的工作参数如表1所示。In this embodiment, the SAR system has two channels (channel 1 and channel 2) working in the strip mode, where channel 1 is used as the reference channel, and the working parameters of the SAR system are shown in Table 1.

载频carrier frequency 5.4GHz5.4GHz 带宽bandwidth 210MHz210MHz 平台速度platform speed 123.47m/s123.47m/s 通道间隔channel spacing 0.612m0.612m 系统PRFSystem PRF 1500Hz1500Hz

表1Table 1

步骤201：对获取的多通道中每一通道的回波数据进行幅度校正和对齐，具体的过程如下：Step 201: Perform amplitude correction and alignment on the acquired echo data of each channel in the multi-channel, the specific process is as follows:

先对获取的多通道中每一通道的回波数据进行幅度校正，包括：对通道2和1的回波数据在方位向上取幅度平均，然后确定通道2与参考通道的幅度平均值之比，根据幅度平均值之比得到通道幅度误差Δa₂(τ)，如图2所示，可以看出这两个通道的距离向的天线方向图在幅度上存在比较大的差异，同时形状也略有不同。First, the amplitude correction is performed on the echo data of each channel in the obtained multi-channel, including: taking the amplitude average of the echo data of channel 2 and 1 in the azimuth direction, and then determining the ratio of the average amplitude of channel 2 to the reference channel, The channel amplitude error Δa ₂ (τ) is obtained according to the ratio of the average amplitude values, as shown in Figure 2, it can be seen that the antenna patterns in the distance direction of the two channels have relatively large differences in amplitude, and their shapes are also slightly different different.

根据公式（3），估计并校正通道幅度误差，然后对校正通道幅度误差后的回拨数据进行对齐，包括：对通道幅度误差校正后的通道2的回波数据和参考通道的回波数据进行距离向FFT，然后将两者共轭相乘，接着在方位向取平均，最后对结果取相位，其结果如图3所示。从图3中可以看出：相位的变化基本是符合线性变化的，距离频率-1×10⁸以下部分和1×10⁸以上部分的剧烈抖动是由于距离向上的过采样而造成的，这些高频信号是雷达带宽以外的噪声。对图3中距离频率在-1×10⁸到1×10⁸之间的相位进行线性拟合，根据拟合直线的斜率确定得到采样起始误差为3.897×10^-9s（秒）。According to the formula (3), estimate and correct the channel amplitude error, and then align the callback data after correcting the channel amplitude error, including: the echo data of channel 2 after the channel amplitude error correction and the echo data of the reference channel FFT in the distance direction, and then multiply the two conjugates, then take the average in the azimuth direction, and finally take the phase of the result, the result is shown in Figure 3. It can be seen from Figure 3 that the change of the phase is basically in line with the linear change, and the severe jitter of the part below the range frequency -1×10 ⁸ and the part above 1×10 ⁸ is caused by the oversampling of the distance. The frequency signal is noise outside the radar bandwidth. Linear fitting is performed on the phase of the distance frequency between -1×10 ⁸ and 1×10 ⁸ in Figure 3, and the sampling start error is determined to be 3.897×10 ^-9 s (seconds) according to the slope of the fitting line.

然后，对对齐后的每一通道的回波数据做距离向IFFT，然后将幅度校正和对齐交替迭代进行；经过三次迭代操作后采样起始误差的估计值为2.642×10^-15s，低于所设定的判决门限1.852×10^-11s，迭代结束。Then, range IFFT is performed on the aligned echo data of each channel, and then the amplitude correction and alignment are alternately iterated; after three iterations, the estimated value of the sampling start error is 2.642×10 ^-15 s, which is lower than The set decision threshold is 1.852×10 ^-11 s, and the iteration ends.

这里，所述判决门限的设定方法为：采样起始误差对应的回波历程相位差2πΔτ_mf₀<0.1π。具体地，通过确定在本例中的判决门限为1.852×10^-11s。Here, the method for setting the decision threshold is as follows: the phase difference of the echo history corresponding to the sampling start error is 2πΔτ _m f ₀ <0.1π. Specifically, by determining that the decision threshold in this example is 1.852×10 ^-11 s.

步骤202：将幅度校正和对齐后的所述回波数据进行成像处理并生成图像，根据所述图像确定代价函数，根据所述代价函数估计通道相位误差，并根据所述通道相位误差对所述回波数据进行校正；具体的过程如下：Step 202: Perform imaging processing on the amplitude-corrected and aligned echo data to generate an image, determine a cost function according to the image, estimate a channel phase error according to the cost function, and calculate the channel phase error according to the channel phase error. The echo data is corrected; the specific process is as follows:

对幅度校正和对齐后的回波数据，采用代价函数对通道相位误差进行估计。图4是本发明实施例中估计通道相位误差的代价函数示意图，如图4所示，未校正的回波数据的代价函数存在峰值A和B，当通道相位误差被完全校正掉后这些峰值都被抑制，由此估计得到的通道相位误差为3.047弧度（rad）。For amplitude-corrected and aligned echo data, a cost function is used to estimate the channel phase error. Fig. 4 is a schematic diagram of a cost function for estimating channel phase error in an embodiment of the present invention. As shown in Fig. 4 , there are peaks A and B in the cost function of uncorrected echo data, and these peaks are all when the channel phase error is completely corrected. is suppressed, the resulting estimated channel phase error is 3.047 radians (rad).

图5为本发明实施例中通道误差校正前成像结果示意图果，图6为本发明实施例中通道误差校正后成像结果示意图，如图5所示，在通道误差校正前图像是模糊的同时虚假目标将原本的真实目标掩盖；如图6所示，通道误差校正之后模糊被抑制，虚假目标消失。Fig. 5 is a schematic diagram of the imaging result before the channel error correction in the embodiment of the present invention, and Fig. 6 is a schematic diagram of the imaging result after the channel error correction in the embodiment of the present invention, as shown in Fig. 5, the image before the channel error correction is blurred and false at the same time The target conceals the original real target; as shown in Figure 6, after channel error correction, the blur is suppressed and the false target disappears.

本发明还提供了一种校正通道误差的装置，如图7所示，所述装置包括第一校正单元71、成像单元72、确定单元73、估计单元74和第二校正单元75；其中，The present invention also provides a device for correcting channel errors. As shown in FIG. 7, the device includes a first correction unit 71, an imaging unit 72, a determination unit 73, an estimation unit 74, and a second correction unit 75; wherein,

所述第一校正单元71，用于对获取的多通道中每一通道的回波数据进行幅度校正和对齐；The first correction unit 71 is configured to perform amplitude correction and alignment on the acquired echo data of each channel in the multi-channel;

所述成像单元72，用于将幅度校正和对齐后的所述回波数据进行成像处理而生成图像；The imaging unit 72 is configured to perform imaging processing on the amplitude-corrected and aligned echo data to generate an image;

所述确定单元73，用于根据所述图像确定代价函数；The determining unit 73 is configured to determine a cost function according to the image;

所述估计单元74，用于根据代价函数估计通道相位误差；The estimation unit 74 is configured to estimate a channel phase error according to a cost function;

所述第二校正单元75，用于根据所述通道相位误差对所述回波数据进行校正。The second correction unit 75 is configured to correct the echo data according to the channel phase error.

进一步的，所述第一校正单元，用于对每一所述通道的所述回波数据依次进行距离向逆快速傅里叶变换以及方位向幅度求平均，根据每一所述通道与参考通道的幅度的平均值之比确定通道幅度误差。Further, the first correction unit is configured to sequentially perform range inverse fast Fourier transform and azimuth amplitude averaging on the echo data of each of the channels, according to each of the channels and the reference channel The ratio of the mean values of the magnitudes determines the channel magnitude error.

进一步的，所述第一校正单元，用于对幅度校正后的每一通道的回波数据和参考信道的回波数据进行距离向快速傅里叶变换，然后将两者共轭相乘，并对共轭相乘后的所述回波数据在方位向求平均，确定求平均后的所述回波数据的相位，根据所述相位的斜率确定通道间的采样起始误差，并对求平均后的所述回波数据与所述相位作积，对所述采样起始误差进行校正。Further, the first correction unit is configured to perform range-to-fast Fourier transform on the amplitude-corrected echo data of each channel and the echo data of the reference channel, and then perform conjugate multiplication of the two, and averaging the echo data after conjugate multiplication in the azimuth direction, determining the phase of the echo data after averaging, determining the sampling start error between channels according to the slope of the phase, and calculating the average The subsequent echo data is multiplied by the phase to correct the sampling start error.

进一步的，所述确定单元，用于将所述图像在方位向进行自相关，并在距离向求平均，将求平均后的结果作为代价函数。Further, the determining unit is configured to perform autocorrelation on the image in the direction of azimuth, and calculate an average in the direction of distance, and use the averaged result as a cost function.

进一步的，所述装置还包括获取单元，所述获取单元用于利用合成孔径雷达系统的条带成像模式获取多通道中每一通道的回波数据。Further, the device further includes an acquisition unit, the acquisition unit is used to acquire the echo data of each channel in the multi-channel by using the strip imaging mode of the synthetic aperture radar system.

本领域的技术人员应当理解，图7所示的校正通道误差的装置中的各处理单元的实现功能可参照前述校正通道误差的相关描述而理解。本领域技术人员应当理解，图7所示的校正通道误差的装置中各处理单元的功能可通过运行于处理器上的程序而实现，也可通过具体的逻辑电路而实现。Those skilled in the art should understand that the functions implemented by each processing unit in the device for correcting channel errors shown in FIG. 7 can be understood with reference to the relevant descriptions for correcting channel errors. Those skilled in the art should understand that the functions of each processing unit in the device for correcting channel errors shown in FIG. 7 can be realized by a program running on a processor, or by a specific logic circuit.

显然，本领域的技术人员应该明白，上述的本发明的各处理单元或各步骤可以用通用的计算装置来实现，其可以集中在单个的计算装置上，或者分布在多个计算装置所组成的网络上，可选地，其可以用计算装置可执行的程序代码来实现，从而，可以将它们存储在存储装置中由计算装置来执行，或者将它们分别制作成各个集成电路模块，或者将它们中的多个模块或步骤制作成单个集成电路模块来实现。这样，本发明不限制于任何特定的硬件和软件结合。Obviously, those skilled in the art should understand that each processing unit or each step of the present invention described above can be implemented by a general-purpose computing device, which can be concentrated on a single computing device, or distributed across a plurality of computing devices. Optionally, it can be implemented with program codes executable by a computing device, thus, they can be stored in a storage device and executed by a computing device, or they can be made into individual integrated circuit modules, or they can be A plurality of modules or steps in the process are realized by making a single integrated circuit module. As such, the present invention is not limited to any specific combination of hardware and software.

以上所述，仅为本发明的较佳实施例而已，并非用于限定本发明的保护范围。The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention.

Claims

1. A method of correcting channel errors, the method comprising:

amplitude correction and alignment are carried out on the acquired echo data of each channel in the multiple channels;

imaging the echo data after amplitude correction and alignment to generate an image, determining a cost function according to the image, estimating a channel phase error according to the cost function, and correcting the echo data according to the channel phase error;

wherein the cost function F_ISC(k) Comprises the following steps:

in the formula, the numerator is the average of autocorrelation, and the function of the denominator is normalization; i (p, q) is a radar image, p and q are respectively a distance coordinate and an azimuth coordinate, k is a coordinate offset, p, q and k are integers, and E is a conjugate_pRepresenting an expectation along the distance direction;

wherein estimating a channel phase error according to a cost function comprises:

for the channel m, the first estimation is performed, firstly, any phase value phi is selected from 0-2 pi_m,1,1Calibrating the echo data of the channel m, imaging the echo data and determining a first cost function; then, at phi_m,1,1Is added with a preset phase value to become phi_m,1,2Imaging the echo data again, determining a second cost function, and comparing curves formed by the first cost function and the second cost function;

if the peak value outside the zero point in the curve formed by the second cost function is suppressed, the fixed phase value is continuously increased until the zero point peak value has the tendency of increasing, and the corrected phase is recorded as phi_m,1,nAnd the true value of the channel phase error is at phi_m,1,n-1To phi_m,1,nTo (c) to (d);

if the zero point peak value in the curve formed by the second cost function has a tendency of becoming larger, the curve indicates that the zero point peak value should be within phi_m,1,1Subtracting a certain phase on the basis to obtain an interval phi_m,1,n-1To phi_m,1,nTo (c) to (d);

in the interval phi_m,1,n-1To phi_m,1,nContinuously repeating the above operations, continuously reducing the added and subtracted preset phase value with the reduction of the interval, and finally determining the channel phase error delta phi of the first estimation of the channel m according to the required estimation precision_m,1；

First estimation of channel phase error Δ φ from channel m_m,1The first channel phase errors of the other channels are obtained in sequence: delta phi_2,1、Δφ_3,1、Δφ_4,1… …, correcting the echo data according to the estimation values, estimating again from channel 2 to obtain the channel phase error delta phi of each channel_2,2、Δφ_3,2、Δφ_4,2… …, respectively; repeating iteration in this way, and stopping iteration when all estimated channel phase errors are smaller than a required threshold value;

in the iterative process of estimating the channel phase error according to the cost function, the phase phi_m,f,gThe meanings of the middle and lower subscripts are as follows: the first index m indicates the channel m, the second index f indicates the f-th estimation, and the third index g indicates the g-th phase taken during said f-th estimation; in particular, the phase phi_m,1,1The first index m in (1) denotes the channel m, the second index 1 denotes the 1 st estimate, and the third index 1Representing the 1 st phase taken during the 1 st estimation; phase phi_m,fThe meanings of the middle and lower subscripts are as follows: the first subscript m denotes the channel m and the second subscript f denotes the f-th estimate; in particular, the phase phi_3,2The first index 3 indicates channel 3 and the second index 2 indicates the phase of the 2 nd estimate.

2. The method of claim 1, wherein the amplitude correcting the echo data of each of the acquired multiple channels comprises:

carrying out azimuth amplitude averaging on the echo data of each channel, and determining a channel amplitude error according to the ratio of the amplitude average value of each channel to a reference channel;

and compensating the echo data of each channel according to the channel amplitude error.

3. The method of claim 1, wherein aligning the echo data for each of the acquired multiple channels comprises:

respectively carrying out distance direction fast Fourier transform on the echo data of each channel after amplitude correction and the echo data of a reference channel, then carrying out conjugate multiplication on the echo data and the echo data, averaging the echo data after conjugate multiplication in the azimuth direction, determining the phase of the echo data after averaging, and determining the sampling initial error between the channels according to the slope of the phase;

and multiplying the echo data of each channel after the amplitude correction with the corresponding phase.

4. A method according to any one of claims 1 to 3, characterized in that the method comprises:

and the amplitude correction and the alignment are alternately and iteratively carried out, and the iteration is finished when the change value of the estimated sampling initial error and/or the channel amplitude error meets the preset judgment threshold.

5. The method of any of claims 1 to 3, wherein determining a cost function from the image comprises:

and carrying out autocorrelation on the image in the azimuth direction, averaging in the distance direction, and taking the averaged result as a cost function.

6. The method of any one of claims 1 to 3, wherein prior to aligning the echo data for each of the acquired multiple channels, the method further comprises:

and acquiring echo data of each channel in multiple channels by utilizing a strip imaging mode of the synthetic aperture radar system.

7. An apparatus for correcting a channel error, the apparatus comprising a first correction unit, an imaging unit, a determination unit, an estimation unit, and a second correction unit; wherein,

the first correction unit is used for performing amplitude correction and alignment on the acquired echo data of each channel in the multiple channels;

the imaging unit is used for imaging the echo data after amplitude correction and alignment to generate an image;

the determining unit is used for determining a cost function according to the image;

the estimating unit is used for estimating a channel phase error according to the cost function;

the second correction unit is used for correcting the echo data according to the channel phase error;

wherein the cost function F_ISC(k) Comprises the following steps:

if the peak value outside the zero point in the curve formed by the second cost function is suppressed, the fixed phase value is continuously increased until the zero point peak valueThere is a tendency to increase, and the corrected phase is recorded as phi_m,1,nAnd the true value of the channel phase error is at phi_m,1,n-1To phi_m,1,nTo (c) to (d);

in the iterative process of estimating the channel phase error according to the cost function, the phase phi_m,f,gThe meanings of the middle and lower subscripts are as follows: the first index m indicates the channel m, the second index f indicates the f-th estimation, and the third index g indicates the g-th phase taken during said f-th estimation; in particular, the phase phi_m,1,1The first index m in (a) denotes the channel m, the second index 1 denotes the 1 st estimation, and the third index 1 denotes the 1 st phase taken during said 1 st estimation; phase phi_m,fThe meanings of the middle and lower subscripts are as follows: the first subscript m denotes the channel m and the second subscript f denotes the f-th estimate; in particular, the phase phi_3,2The first index 3 indicates channel 3 and the second index 2 indicates the phase of the 2 nd estimate.

8. The apparatus of claim 7, wherein the first correction unit is further configured to: carrying out azimuth amplitude averaging on the echo data of each channel, and determining a channel amplitude error according to the ratio of the average value of the amplitudes of each channel and a reference channel; and compensating the echo data of each channel according to the channel amplitude error.

9. The apparatus of claim 7, wherein the first correction unit is further configured to: carrying out fast Fourier transform on the distance direction of echo data of each channel after amplitude correction and echo data of a reference channel, then carrying out conjugate multiplication on the echo data and the echo data, averaging the echo data after conjugate multiplication in the azimuth direction, determining the phase of the echo data after averaging, and determining the sampling initial error between the channels according to the slope of the phase; and multiplying the echo data of each channel after the amplitude correction with the corresponding phase.

10. The apparatus according to any of claims 7 to 9, wherein the determining unit is further configured to: and carrying out autocorrelation on the image in the azimuth direction, averaging in the distance direction, and taking the averaged result as a cost function.