Radar imaging method based on step frequency dechirp and adaptive phase compensation
By introducing step frequency de-absorbing and adaptive phase compensation technology into the radar system, the problem of signal frequency inconsistency caused by target motion is solved, high-resolution images and low sampling rate radar imaging is realized, and high-precision motion compensation is met need.
Patent Information
- Application Number
- CN202411071275.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-06
- Publication Date
- 2024-11-15
AI Technical Summary
In radar systems, the target motion causes the desloping reference signal delay to be unable to accurately respond, resulting in inconsistent signal frequencies between different sub-pulses, unable to achieve phase-parallel accumulation, and the system sampling rate is too high or insufficient bandwidth.
The radar imaging method based on step frequency de-absorption and adaptive phase compensation is adopted. The phase error between sub-pulses is compensated through bandwidth coherence processing, and the de-absorption reference signal is kept unchanged, and the frequency domain envelope alignment and initial phase correction are performed. Ensure the signal-to-noise ratio and accuracy of one-dimensional distance image.
The system sampling rate and calculation amount are reduced, the motion compensation accuracy is improved, the high-resolution image and signal-to-noise ratio are ensured, and the conditions are provided for target recognition.
Smart Images

Figure CN118962675A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to radar signal processing, and in particular to a radar imaging method based on stepped frequency de-skewing and adaptive phase compensation. Background Art
[0002] In order to obtain more abundant target information, modern radars usually use large bandwidth signals. If large bandwidth linear frequency modulation signals are directly sampled and pulse compressed, high requirements are placed on the A / D sampling rate and the processing capacity of the signal processing unit. In order to solve the problem that the system sampling rate is still too high after de-sloping or that a sufficiently large bandwidth signal cannot be generated due to technical limitations, the concept of frequency stepping is introduced in the de-sloped linear frequency modulation system. Through signal processing, a few sub-pulses are used to synthesize a sufficiently large bandwidth, and the system sampling rate after de-sloping is kept low enough.
[0003] In actual systems, the target is usually moving. When the target moves, the system inevitably has tracking errors, so that the delay of the de-skewed reference signal cannot accurately reflect the motion state of the target, resulting in the delay difference between the two no longer remaining constant between different sub-pulses. Therefore, the signal frequency of the same scattering point in different sub-pulse de-skew echoes is inconsistent, and coherent accumulation cannot be achieved. Summary of the invention
[0004] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a radar imaging method based on stepped frequency de-skewing and adaptive phase compensation. Drawing on the concept of bandwidth coherent processing, the bandwidth coherent processing method is used to compensate for the phase error between sub-pulses, so as to ensure the coherence between sub-band pulses with a lower amount of calculation when the motion compensation accuracy requirement exceeds the accuracy range of the inertial navigation system, thereby ensuring the signal-to-noise ratio and accuracy of the one-dimensional range image.
[0005] The object of the present invention is achieved through the following technical scheme: A radar detection method based on stepped frequency de-skewing and bandwidth synthesis, comprising the following steps:
[0006] S1. Processing the radar echo signal based on frequency modulation de-skewing;
[0007] S2. In the step frequency signal, the time delay of the de-skew reference signal is kept unchanged in the step frequency signal to compensate for the phase error caused by the target speed;
[0008] S3. Convert each sub-pulse into the frequency domain dimension through FFT transformation and perform frequency domain envelope alignment;
[0009] S4. Align the echoes of targets at different distances in terms of distance, and implement de-skewing and residual phase removal processing;
[0010] S5. Taking the first sub-pulse as a reference, an initial phase value is estimated, and the echo of each sub-pulse is compensated respectively to achieve coherent initial phase correction;
[0011] S6. Splice the sub-pulse signals in the time domain and then perform FFT to obtain a high-resolution one-dimensional range image.
[0012] The beneficial effects of the present invention are as follows: the present invention introduces the concept of frequency stepping in the de-sloped linear frequency modulation system, which is equivalent to extending the duration of a signal with the same bandwidth and transmitting it in batches, and then synthesizing it by signal processing, using a few sub-pulses to synthesize a sufficiently large bandwidth, and keeping the system sampling rate after de-sloping sufficiently low, thereby reducing the requirements of the broadband system on the A / D sampling rate and signal processing, and at the same time, the large bandwidth ensures the resolution of the image, providing conditions for subsequent target recognition;
[0013] The phase between sub-bands is compensated by coherence of spectral overlap, and the coherence between sub-band pulses is guaranteed with low computational complexity when the motion compensation accuracy requirement exceeds the accuracy range of the inertial navigation system, thereby ensuring the signal-to-noise ratio and accuracy of the one-dimensional range image. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] Figure 1 The present invention is a flow chart of the method. DETAILED DESCRIPTION
[0015] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings, but the protection scope of the present invention is not limited to the following.
[0016] like Figure 1 As shown, a radar detection method based on stepped frequency de-skewing and bandwidth synthesis includes the following steps:
[0017] S1. Processing the radar echo signal based on frequency modulation de-skewing;
[0018] S101. First, analyze the FM de-skew technology: The de-skew technology is to mix the received echo signal with a linear FM reference signal with the same FM slope, and convert the signal whose frequency changes linearly with time into a single frequency signal. The reference signal is symmetrical about the center distance of the scene and has a modulation rate opposite to that of the transmitted pulse.
[0019] The pulse signal is transmitted in sequence with a repetition period T, that is, the transmission time t n = nT, called slow time, n represents the sub-pulse number, and the time starting from the emission time of the current sub-pulse is expressed as represents, called fast time, t represents full time, The linear frequency modulation signal is expressed as:
[0020]
[0021] Where: T p is the signal pulse width; f c is the center frequency; γ is the signal modulation frequency;
[0022] Let the reference distance be R ref , then the reference signal is:
[0023]
[0024] Where, T ref Indicates the pulse width of the reference signal and is greater than T p ;
[0025] Assume that the distance from a target to the radar is R t , the radar receives the target's echo signal for:
[0026]
[0027] A is the amplitude of the received echo signal. If R Δ =R t -R ref , then the output signal after de-skewing is:
[0028]
[0029] Echo frequency after de-skewing f i , all echoes become single-frequency pulse signals, and the pulse width is still T p , and the carrier frequency value is related to R Δ Proportional, Perform Fourier transform on the de-skewing signal and obtain the corresponding sinc-shaped narrow pulses of each echo in the frequency domain, with a resolution of 1 / T p , the distance resolution is
[0030] S102. Extend the de-skewing process to the processing of the frequency-modulated step signal: Assume that the number of steps detected by the radar is N, and the time interval between each sub-pulse is T r , the starting frequency is f0, the bandwidth of the signal is B, the frequency modulation slope is γ, and the pulse duration is T p , the center frequency f0 of the sub-pulse signal will increase with a fixed frequency value Δf, covering the bandwidth of the full aperture. The mathematical model of the frequency-modulated step signal is:
[0031]
[0032] In the formula,
[0033] f n=f0+(n-1)Δf,n=1,…,N; 0≤[t-nT r ]≤T p ;
[0034] S103. The echo signal of the emitted step frequency modulation signal after being reflected by a point target at a distance R from the radar is
[0035]
[0036] In the formula, τ = 2R / c corresponds to the target distance delay;
[0037] S104. Select the delay τ0 = 2R0 / c, the time width is T ref The signal is the de-skewed reference signal:
[0038]
[0039] The signal after de-skewing and low-pass filtering is:
[0040]
[0041] S2. In the step frequency signal, the time delay of the de-skew reference signal is kept unchanged in the step frequency signal to compensate for the phase error caused by the target speed;
[0042] S201. In actual systems, the target is usually moving. When the target moves, there is inevitably a tracking error in the system. In order to make the sub-pulse de-skewed echo signal of the moving target re-coherent, the time delay of the de-skewed reference signal can be kept unchanged within a set of stepped frequency signals to compensate for the phase error caused by the target speed. Assuming that the radial speed of the target is v, the distance error caused by the speed is:
[0043] R en =v(n-1)T r ,n=1,…N
[0044] S202. R en Substituting the output signal model after de-skewing in S101 and ignoring the quadratic term, the phase error caused by the distance error caused by the speed can be obtained. In order to compensate for the phase error, the phase error compensation function caused by the target speed is calculated (ignoring the quadratic term):
[0045]
[0046] S203. Assuming that the relative positions of the target points in the target scene remain unchanged during bandwidth synthesis, the sub-pulse de-slanted echo signal is respectively combined with the above formula s according to the pulse sequence number. cn The corresponding multiplication of (t) can compensate for the phase error caused by the target speed, and the echo of each point target becomes a single-frequency echo.
[0047] S3. Convert each sub-pulse into the frequency domain dimension through FFT transformation and perform frequency domain envelope alignment;
[0048] The step S3 comprises:
[0049] S301. Convert each sub-pulse into the frequency domain by performing an FFT change on the signal after compensation in S203;
[0050]
[0051] Among the three phase terms, the first one is the Doppler term, and the last two are the terms to be removed;
[0052] S302. Perform frequency domain envelope alignment:
[0053] Taking the first sub-pulse echo as the reference, the time domain signal of each sub-pulse is time-shifted. The time shift of the nth sub-pulse is
[0054] Δt n =-2*v(n-1)T r / c
[0055] Through the time shift characteristics of Fourier transform, the time shift S of the nth sub-pulse is ali_n (f) By using the frequency domain signal S if (f i ) multiplied by the time shift coefficient S ali_n , and obtain the frequency domain signal after the envelopes of each sub-pulse are aligned, where:
[0056] S ali_n (f) = exp(-j2πfΔt n ), n=1,…N.
[0057] S4. Align the echoes of targets at different distances in terms of distance to achieve de-skewing processing;
[0058] The echoes at different distances are staggered in time, which is called skewing. This time staggering does not bring new information, but brings inconvenience to the subsequent processing. Therefore, it is usually hoped that the echoes of targets at different distances will be aligned in distance, which is called de-skewing. As a result of de-skewing, the RVP item disappears.
[0059] The second phase in step S301 is the residual phase term, i.e., the RVP term. The third phase term indicates that the echoes at different distances are staggered in time, which is called the envelope tilt term. However, this time stagger does not bring new information. Therefore, when the echoes of targets at different distances are aligned in distance, the RVP term and the envelope tilt term should be removed.
[0060] The second and third terms are both in phase with R ΔWhen the difference frequency echo changes to the difference frequency domain, it becomes a sinc function, and its peak value is located at For distance R Δ The goal is to compensate, as long as the compensation The phase at
[0061] S if The last two phases of are written as:
[0062]
[0063] Therefore, the frequency domain signal after the envelope alignment of each sub-pulse is multiplied by the compensation factor of the following formula, and then returned to the time domain through inverse transformation to complete the compensation. The two phase terms of RVP and envelope tilt term are removed to obtain the time domain signal after compensation of each sub-pulse; the compensation factor is recorded as S c
[0064]
[0065] S5. Taking the first sub-pulse as a reference, an initial phase value is estimated, and the echo of each sub-pulse is compensated respectively to achieve coherent initial phase correction;
[0066] S501. When there is noise and speed compensation error, the phase will be discontinuous, and the synthesized one-dimensional range image will have increased side lobes. Considering that the transition bands on both sides of the band will cause the side lobes to increase, the overlapping bandwidth is designed to be more than 10% of the total bandwidth in the radar system, and the output signal is brought into S101 after de-skewing, completing the compensation of the RVP term and the envelope skew term and considering the residual motion error, rewritten as
[0067]
[0068] in v re The residual motion error is caused by the insufficient accuracy of velocity compensation. The azimuth phase error correction is realized by overlapping frequency bands, which is directly processed in the time domain. However, due to the influence of the residual motion of the target, the phase between adjacent sub-pulses jumps, which seriously affects the sub-pulse splicing and the synthesis of high-resolution one-dimensional range images. Therefore, it is necessary to correct the phase error between adjacent sub-pulses.
[0069] S502. The phase coherence between adjacent sub-pulses is achieved by using the conjugate phase difference method. The overlapping parts of two adjacent pulses are respectively subjected to Fourier transform, and the conjugate multiplication is used to obtain the phase
[0070]
[0071] Taking the first sub-pulse in the pulse group as a reference, the above formula can be used to estimate the N-1 phase differences between all N sub-pulses in the pulse group. By adding them together, the phase factor caused by the nth residual velocity can be obtained.
[0072]
[0073] The initial phase is compensated by the phase factor, and each compensated time-domain sub-pulse is multiplied by the corresponding phase compensation factor, so that the phases between the sub-pulses will be continuous.
[0074] S6. Splice the sub-pulse signals in the time domain and then perform FFT to obtain a high-resolution one-dimensional range image.
[0075] The de-slant echo of the next sub-pulse in the step pulse train has an additional de-slant echo of Δf / γ Determine the time offset.
[0076] S601. Select Δf and sampling rate fs to satisfy T = Δf / γ , and fs·T is an integer, so that the de-oblique echo phase of the next sub-pulse is completely continuous with the de-oblique echo phase of the previous sub-pulse;
[0077] S602. Convert each sub-pulse into a time domain signal through ifft, and then select the parameters of the frequency modulation step signal. At this time, each sub-pulse is time-domain aligned. The number of points selected for each sub-pulse is The data is spliced in the order of time-stepping frequency to form a signal with large time width and bandwidth;
[0078] S603. Perform Fft on the spliced large bandwidth signal to obtain a high-resolution one-dimensional range image.
[0079] The above is a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the form disclosed herein, and should not be regarded as excluding other embodiments, but can be used in other combinations, modifications and environments, and can be modified within the scope of the concept described herein through the above teachings or the technology or knowledge of the relevant field. The changes and modifications made by those skilled in the art do not depart from the spirit and scope of the present invention, and should be within the scope of protection of the claims attached to the present invention.
Claims
1. A radar detection method based on stepped frequency de-skewing and bandwidth synthesis, characterized in that: The following steps are involved: S1. Processing the radar echo signal based on frequency modulation de-skewing; S2. In the step frequency signal, the time delay of the de-skew reference signal is kept unchanged in the step frequency signal to compensate for the phase error caused by the target speed; S3. Convert each sub-pulse into the frequency domain dimension through FFT transformation and perform frequency domain envelope alignment; S4. Align the echoes of targets at different distances in terms of distance, and implement de-skewing and residual phase removal processing; S5. Taking the first sub-pulse as a reference, an initial phase value is estimated, and the echo of each sub-pulse is compensated respectively to achieve coherent initial phase correction; S6. Splice the sub-pulse signals in the time domain and then perform FFT to obtain a high-resolution one-dimensional range image.
2. The radar imaging method based on stepped frequency de-skewing and adaptive phase compensation according to claim 1, characterized in that: The step S1 comprises: S101. Perform frequency modulation de-skewing on the radar echo signal: frequency modulation de-skewing is to mix the received echo signal with a linear frequency modulation reference signal with the same frequency modulation slope, so as to convert the signal whose frequency changes linearly with time into a single frequency signal, and the reference signal is symmetrical about the center distance of the scene and has a frequency modulation opposite to that of the transmitted pulse; The pulse signal is transmitted in sequence with a repetition period T, that is, the transmission time t n = nT, called slow time, n represents the sub-pulse number, and the time starting from the emission time of the current sub-pulse is expressed as represents, called fast time, t represents full time, The linear frequency modulation signal is expressed as: Where: T p is the signal pulse width; f c is the center frequency; γ is the signal modulation frequency; Let the reference distance be R ref , then the reference signal is: Where, T ref Indicates the pulse width of the reference signal and is greater than T p ; Assume that the distance from a target to the radar is R t , the radar receives the target's echo signal for: A is the amplitude of the received echo signal. If R Δ =R t -R ref , then the output signal after de-skewing is: Echo frequency after de-skewing f i , all echoes become single-frequency pulse signals, and the pulse width is still T p , and the carrier frequency value is related to R Δ Proportional, Perform Fourier transform on the de-skewing signal and obtain the corresponding sinc-shaped narrow pulses of each echo in the frequency domain, with a resolution of 1 / T p , the distance resolution is S102. Extend the de-skewing process to the processing of the frequency-modulated step signal: Assume that the number of steps detected by the radar is N, and the time interval between each sub-pulse is T r , the starting frequency is f0, the bandwidth of the signal is B, the frequency modulation slope is γ, and the pulse duration is T p , the center frequency f0 of the sub-pulse signal will increase with a fixed frequency value Δf, covering the bandwidth of the full aperture. The mathematical model of the frequency-modulated step signal is: In the formula, f n =f0+(n-1)Δf,n=1,…,N;0≤[t-nT r ]≤T p ; S103. The echo signal of the transmitted step frequency modulation signal after being reflected by a point target at a distance R from the radar is In the formula, τ = 2R / c corresponds to the target distance delay; S104. Select the delay τ0 = 2R0 / c, the time width is T ref The signal is the de-skewed reference signal: The signal after de-skewing and low-pass filtering is:
3. The radar imaging method based on stepped frequency de-skewing and adaptive phase compensation according to claim 1, characterized in that: The step S2 comprises: The step S2 comprises: S201. Assume that the time interval between each sub-pulse is T r , the target radial velocity corresponding to the nth sub-pulse is v(n), and the distance error caused by the velocity is calculated as: R en =v(n-1)T r ,n=1,…N; S202. R en Substitute the output signal model after de-skewing in S101 and ignore the quadratic term to obtain the phase error caused by the distance error caused by the speed. In order to compensate for the phase error, the phase error compensation function caused by the target speed is calculated as: S203. Assuming that the relative positions of the target points in the target scene remain unchanged during bandwidth synthesis, the sub-pulse de-slanted echo signals are respectively combined with s cn (t) are multiplied correspondingly to compensate for the phase error caused by the target speed, and the echo of each point target becomes a single-frequency echo.
4. The radar imaging method based on stepped frequency de-skewing and adaptive phase compensation according to claim 3, characterized in that: The step S3 comprises: The step S3 comprises: S301. Convert each sub-pulse into the frequency domain by performing an FFT change on the signal after compensation in S203; Among the three phase terms, the first one is the Doppler term, and the last two are the terms to be removed; S302. Perform frequency domain envelope alignment: Taking the first sub-pulse echo as the reference, the time domain signal of each sub-pulse is time-shifted. The time shift of the nth sub-pulse is Δt n =-2*v(n-1)T r / c Through the time shift characteristics of Fourier transform, the time shift S of the nth sub-pulse is ali_n (f) By using the frequency domain signal S if (f i ) multiplied by the time shift coefficient S ali_n , and obtain the frequency domain signal after the envelopes of each sub-pulse are aligned, where: S ali_n (f)=exp(-j2πfΔt n ),n=1,…N。 5. The radar imaging method based on stepped frequency de-skewing and adaptive phase compensation according to claim 1, characterized in that: The step S4 comprises: The second phase in step S301 is the residual phase term, i.e., the RVP term. The third phase term indicates that the echoes at different distances are staggered in time, which is called the envelope tilt term. When the echoes of targets at different distances are aligned in distance, the RVP term and the envelope tilt term should be removed. The second and third terms are both in phase with R Δ When the difference frequency echo changes to the difference frequency domain, it becomes a sinc function, and its peak value is located at For distance R Δ The goal is to compensate, as long as the compensation The phase at S if The last two phases of are written as: Therefore, the frequency domain signal after the envelope alignment of each sub-pulse is multiplied by the compensation factor of the following formula, and then returned to the time domain through inverse transformation to complete the compensation. The two phase terms of RVP and envelope tilt term are removed to obtain the time domain signal after compensation of each sub-pulse; the compensation factor is recorded as S c 6. The radar imaging method based on stepped frequency de-skewing and adaptive phase compensation according to claim 1, characterized in that: The step S5 comprises: S501. The overlapping bandwidth is designed to be more than 10% of the total bandwidth. According to the output signal after the de-skewing process in S101, the compensation of the RVP term and the envelope skewing term is completed and the residual motion error is considered, which is rewritten as in v re The residual motion error is caused by the insufficient accuracy of velocity compensation. The azimuth phase error correction is realized by overlapping frequency bands and processed directly in the time domain. However, due to the influence of the residual motion of the target, the phase between adjacent sub-pulses jumps. Therefore, the phase error between adjacent sub-pulses needs to be corrected. S502. The phase coherence between adjacent sub-pulses is achieved by using the conjugate phase difference method. The overlapping parts of two adjacent pulses are respectively subjected to Fourier transform, and the conjugate multiplication is used to obtain the phase: Taking the first sub-pulse in the pulse group as a reference, the above formula is used to estimate the N-1 phase differences between all N sub-pulses in the pulse group, and the phase factor caused by the nth residual velocity is obtained through cumulative operation; The initial phase is compensated by the phase factor, and each compensated time-domain sub-pulse is multiplied by the corresponding phase compensation factor, so that the phases between the sub-pulses will be continuous.
7. The radar imaging method based on stepped frequency de-skewing and adaptive phase compensation according to claim 1, characterized in that: The step S6 comprises: S601. Select Δf and sampling rate fs to satisfy T=Δf / γ , and fs·T is an integer, so that the de-oblique echo phase of the next sub-pulse is completely continuous with the de-oblique echo phase of the previous sub-pulse; S602. Convert each sub-pulse into a time domain signal through ifft, and then select the parameters of the frequency modulation step signal. At this time, each sub-pulse is time-domain aligned, and the number of points selected for each sub-pulse is The data is spliced in the order of time-stepping frequency to form a signal with large time width and bandwidth; S603. Perform Fft on the spliced large bandwidth signal to obtain a high-resolution one-dimensional range image.
Citation Information
Cited By
Slow moving object detection method, device and equipment based on security radar
CN119199794A
