A doppler migration compensation method based on phase compensation and front pulse feedback

Abstract

The application discloses a Doppler migration compensation method based on phase compensation and front pulse feedback, which comprises the following steps: firstly, performing Hilbert transform and complex conjugate processing on a local reference signal, mixing the local reference signal with an original echo signal, and obtaining a mixed signal of a compressed pulse and extracted Doppler frequency shift; calculating a peak frequency of the mixed signal and saving it; then, during non-initial section accumulation, constructing a phase compensation function according to the Doppler frequency shift obtained from the last section signal, performing phase compensation on the mixed signal and calculating a peak frequency, performing a cyclic shift operation on a historical accumulated spectrum according to the estimation result, and compensating the Doppler frequency shift caused cross-spectrum unit migration; finally, performing non-coherent accumulation on the spectrum data after cyclic alignment. The application can estimate the number of Doppler migration units caused by target speed change without searching for motion parameters, compensate the Doppler migration effect in non-coherent accumulation, has low calculation amount, strong real-time performance, and avoids the loss of non-coherent accumulation gain of the signal.

Description

Technical Field

[0001] This invention relates to the field of coherent lidar signal processing, and in particular to a Doppler migration compensation method based on phase compensation and front-end pulse feedback. Background Technology

[0002] With the rapid development of fields such as autonomous driving, aerospace, and deep space exploration, lidar has also seen rapid development and application due to its high detection sensitivity and strong anti-interference capabilities. The detection of highly dynamic, long-range, and weak targets has become an important research direction for coherent lidar systems. In lidar systems, to achieve effective detection of long-range, weak targets, signal energy accumulation techniques such as noncoherent accumulation are typically employed to improve the signal-to-noise ratio of the echo signal, thereby calculating information such as the target's distance and velocity.

[0003] However, in real-world high-dynamic applications, targets often exhibit high-order motion (such as high acceleration or variable acceleration). This complex motion characteristic causes the Doppler frequency shift of the echo signal to change rapidly and non-linearly over time, with the target Doppler migration easily exceeding the width of a single frequency domain resolution cell. This phenomenon is known as Doppler migration in the field of radar signal processing.

[0004] The Doppler migration phenomenon severely impacts the gain of energy accumulation. Specifically, it causes signal energy that should be concentrated in a single frequency cell to spread across and disperse into multiple adjacent frequency cells over time, resulting in significant spectral broadening. This energy dispersion not only leads to a sharp decrease in the signal-to-noise ratio gain of noncoherent accumulation, resulting in increased false alarm and missed alarm rates, but also significantly reduces the accuracy of Doppler center frequency extraction, thereby decreasing the accuracy of the system's target motion estimation.

[0005] Although some compensation algorithms based on parameter search (such as the generalized Radon Fourier transform and fractional Fourier transform) exist, these algorithms typically suffer from long search times and high computational complexity, making it difficult to meet the real-time requirements of practical engineering. Therefore, how to simply and quickly estimate the number of Doppler migration elements, effectively correct Doppler migration, and ensure cumulative gain in low signal-to-noise ratio environments when the target undergoes high-order motion that induces Doppler migration has become an urgent problem to be solved in the field of coherent lidar signal processing. Summary of the Invention

[0006] In view of the aforementioned shortcomings of existing technologies, the technical problem to be solved by this invention is the poor real-time performance and high computational complexity of existing Doppler migration compensation techniques. This invention provides a Doppler migration compensation method based on phase compensation and front-end pulse feedback. It can estimate the number of Doppler migration units caused by changes in target velocity without searching for motion parameters, and then compensate for the Doppler migration effect in noncoherent accumulation. Furthermore, the algorithm is simple, computationally inexpensive, and can effectively improve the real-time performance of signal processing while avoiding the loss of signal noncoherent accumulation gain.

[0007] To achieve the above objectives, the present invention provides a Doppler migration compensation method based on phase compensation and front-end pulse feedback, comprising the following steps:

[0008] Step S1: Perform Hilbert transform on the local signal to construct a complex signal, and take its conjugate to obtain the local complex conjugate signal;

[0009] Step S2: Mix the local complex conjugate signal with the received echo signal to obtain a mixed signal with pulse compression and Doppler frequency shift extraction.

[0010] Step S3: Calculate the peak frequency of the spectrum of the mixing signal, i.e., the Doppler frequency shift, and save the Doppler frequency shift;

[0011] Step S4: Determine whether the signal being processed is the first segment of the current accumulation block. If yes, perform non-coherent accumulation on the spectrum of the current signal directly. If no, proceed to steps S5 to S8.

[0012] Step S5: Extract the Doppler frequency shift saved from the previous mixing signal to construct the phase compensation function;

[0013] Step S6: Apply the phase compensation function to perform phase compensation on the mixing signal, and calculate the peak frequency of the spectrum of the phase-compensated signal, which is the number of Doppler migration units of the target.

[0014] Step S7: Cyclicly shift the historical cumulative spectrum according to the number of Doppler migration units of the target to compensate for Doppler migration;

[0015] Step S8: Finally, the compensated historical cumulative spectrum is non-coherently accumulated with the current segment mixing signal spectrum.

[0016] Further, the local signal in step S1 is represented as ,in It is the local signal amplitude. It's phase noise. The local frequency is used to construct a complex signal by performing a Hilbert transform on the local signal, and then taking its conjugate to obtain the local complex conjugate signal. .

[0017] Further, in step S2, the local complex conjugate signal is mixed with the received echo signal to obtain a mixed signal with pulse compression and Doppler frequency shift extraction, specifically including:

[0018] The local complexized conjugate signal is represented as ,in It is the local signal amplitude. It is a local frequency. It is the phase noise of the signal, and the echo signal is represented as... ,in It is the echo amplitude. It's a Doppler shift. It is signal noise. Multiplying the local complex conjugate signal with the echo signal and mixing them, the resulting mixed signal is: ,in It is the amplitude of the mixing signal. It is the noise cross term in the mixing process, that is, the noise of the mixing signal.

[0019] Furthermore, Doppler shift refers to the reference Doppler shift value used for compensation in the next data segment. .

[0020] Further, in step S5, the Doppler frequency shift saved from the previous mixing signal is extracted to construct a phase compensation function, the mathematical expression of which is: ,in, For phase compensation function, It is the imaginary unit.

[0021] Further, in step S6, a phase compensation function is applied to perform phase compensation on the mixing signal, and the number of Doppler migration units of the target is calculated. Specifically, this includes: applying the phase compensation function... Multiply the current segment's mixing signal by a complex number to obtain the compensated mixing signal;

[0022] Perform a Fourier transform on the compensated mixed signal to obtain the compensated amplitude spectrum, and search for the peak index positions of this compensated amplitude spectrum starting from 0. ;

[0023] Calculate the number of Doppler migrating units based on the peak index position. .

[0024] Furthermore, the number of Doppler migrating units is calculated based on the peak index position. The calculation rule is as follows: when hour, ;when hour, ,in This represents the number of sampling points for a single signal segment.

[0025] Furthermore, in step S7, the historical cumulative spectrum is cyclically shifted according to the number of Doppler migrating units of the target. The cyclic shifting method is to move towards the latest target Doppler frequency point based on the estimated number of target Doppler migrating units.

[0026] Furthermore, the cyclic shift method is specifically as follows: If the value is greater than 0, then shift to the right in a circular fashion. Bit, If the value is less than 0, then shift to the left in a circular shift. Bit.

[0027] Technical effect

[0028] (1) The present invention provides a Doppler migration compensation method based on phase compensation and front-end pulse feedback. A phase compensation function is constructed based on the Doppler frequency shift information obtained from the front-end pulse data, and then the current pulse information is phase-compensated. A signal containing only the Doppler migration frequency is constructed. The number of Doppler migration units between two adjacent pulse segments is obtained by calculating the peak frequency index of this signal. Based on the calculated number of Doppler migration units, the accumulated spectrum is cyclically shifted to compensate for Doppler migration, so that energy is focused at the latest Doppler frequency point. This method can effectively overcome the non-coherent accumulation spectrum broadening phenomenon caused by the Doppler transformation of maneuvering targets, improve the non-coherent accumulation gain, effectively suppress Doppler migration, and achieve energy focusing.

[0029] (2) The method of the present invention only requires one fast Fourier transform and M complex multiplications (complex multiplication in step S6) to estimate the number of Doppler migrating units, where M refers to the length of each data segment. The algorithm has low complexity. Furthermore, the method does not require complex motion parameter search to estimate the number of Doppler migrating units, and the algorithm has strong real-time performance and low computational complexity.

[0030] The following will further explain the concept, specific structure, and technical effects of the present invention in conjunction with the accompanying drawings, so as to fully understand the purpose, features, and effects of the present invention. Attached Figure Description

[0031] Figure 1 This is a flowchart of a Doppler migration compensation method based on phase compensation and front-end pulse feedback according to the present invention;

[0032] Figure 2 This is a simulation diagram of echo theory Doppler frequency shift variation of a Doppler migration compensation method based on phase compensation and front-end pulse feedback according to the present invention;

[0033] Figure 3This is a comparison chart of the direct accumulation spectrum and the compensated Doppler migration accumulation spectrum after 8-segment accumulation of a Doppler migration compensation method based on phase compensation and front-end pulse feedback according to the present invention.

[0034] Figure 4 This is a comparison chart of the direct cumulative signal-to-noise ratio and the compensated cumulative signal-to-noise ratio of 20 sets of data from a Doppler migration compensation method based on phase compensation and front-end pulse feedback according to the present invention. Detailed Implementation

[0035] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0036] In the following description, specific details, such as particular internal procedures and techniques, are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will appreciate that the invention may be practiced in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of the invention with unnecessary detail.

[0037] This invention provides a Doppler migration compensation method based on phase compensation and front-end pulse feedback. First, the local reference signal undergoes Hilbert transform and complex conjugation processing, and is then mixed with the original echo signal to obtain a pulse-compressed and Doppler-shifted mixed signal. The peak frequency of the mixed signal, i.e., the Doppler shift, is calculated and the result is saved for feedback to the next pulse data segment. Then, during non-initial accumulation, a phase compensation function is constructed based on the Doppler shift obtained from the previous signal segment to perform phase compensation on the mixed signal. The peak frequency of the phase-compensated mixed signal, i.e., the number of Doppler migration units of the target, is calculated. Based on the estimation result, a cyclic shift operation is performed on the historical accumulated spectrum to compensate for the cross-spectral unit movement caused by the Doppler shift. Finally, non-coherent accumulation is performed on the cyclically aligned spectrum data. This invention can estimate the number of Doppler migration units caused by target velocity changes without searching for motion parameters, and then compensate for the Doppler migration effect in non-coherent accumulation. Furthermore, the algorithm is simple, computationally inexpensive, and can effectively improve the real-time performance of signal processing while avoiding the loss of signal gain in non-coherent accumulation.

[0038] like Figure 1 As shown, this embodiment of the invention provides a Doppler migration compensation method based on phase compensation and front-end pulse feedback, specifically including the following steps:

[0039] S1, for coherent lidar, its local signal is represented as: ,in It is the local signal amplitude. It's phase noise. It is a local frequency. In this embodiment of the invention, the local signal is first... Perform a Hilbert transform to extract its orthogonal components. Subsequently, using the original signal as the real part and the orthogonal components as the imaginary part, it is complexified to construct a complex signal (i.e., an analytic signal), thus obtaining the complex signal. According to Euler's formula, this complex signal can be simplified to exponential form. Finally, the complex signal By taking the complex conjugate, it is possible to construct a local complex conjugate signal in which both the local frequency and phase noise have negative signs. .

[0040] S2, the echo signal of the coherent lidar is represented as (in It is the echo amplitude. It's a Doppler shift. (This is signal noise). Multiplying the local complexized conjugate signal obtained in step S1 with the received echo signal in the time domain yields the following mixed signal: ,in It is the amplitude of the mixing signal. This refers to the noise cross-term in the mixing process, i.e., the noise in the mixed signal. It can be seen that this method compensates for the phase noise of the signal, reduces the spectral broadening of the echo signal, and compresses the pulse to exceed the amplitude of the noise. This is also a prerequisite for subsequently constructing a signal containing only the Doppler migration frequency. Furthermore, the signal mixing compensates for the local frequency, and the signal retains only the Doppler frequency shift.

[0041] S3, the mixing signal obtained from S2 Perform a Fourier transform to obtain the signal spectrum, and then determine the peak frequency of the spectrum, which is the Doppler frequency shift of that segment of the signal. Save it for the construction of the phase compensation function in the next pulse segment.

[0042] S4, Extract the segment index of the currently processed signal. (in =1, 2, ..., , (The total number of segments participating in each accumulation), determined by the segment index. Whether the current signal is equal to 1 determines whether it is the first segment of the signal: if yes, then the spectrum of the current signal is directly accumulated non-coherently, that is, the spectrum of the first accumulated signal is used as the historical accumulated spectrum; if no, then proceed to the next step.

[0043] S5, extract the Doppler frequency shift saved from the previous mixing signal to construct the phase compensation function, the specific mathematical expression of which is as follows: ,in, The phase compensation function is... It is the imaginary unit.

[0044] S6, assuming that due to the target's acceleration, jerk, and other high-order complex motions, a Doppler change occurs between two adjacent signal segments. The change in Doppler frequency reflects the difference in Doppler frequency between the current signal segment and the previous signal segment. If the frequency unit migrates, then the mixing signal of the current segment can be expressed as:

[0045] to By multiplying in the time domain and performing phase compensation on the mixed signal, the signal can be obtained. The index corresponding to the peak frequency of the signal's spectrum, i.e., the number of Doppler migration units between adjacent segments. The calculation rule is as follows: when hour, ;when hour, ,in This represents the number of sampling points for a single signal segment.

[0046] S7, based on the obtained number of Doppler migration units of the target, performs a cyclic shift on the historical cumulative spectrum to compensate for Doppler migration. The cyclic shift method here is as follows: If the value is greater than 0, then shift to the right in a circular fashion. Bit, If the value is less than 0, then shift to the left in a circular shift. This is a cyclic shifting method that moves towards the latest target Doppler frequency based on the estimated number of target Doppler migrating units.

[0047] S8. Finally, the compensated historical accumulated spectrum is non-coherently accumulated with the current segment's mixing signal spectrum. The non-coherent accumulation method is to directly accumulate the corresponding amplitude spectra. Then, the peak frequency of the accumulated spectrum is calculated, which is the target's current Doppler frequency shift. The target velocity is then calculated using the Doppler formula.

[0048] To demonstrate the effectiveness of the algorithm, this embodiment uses Matlab to simulate it. Under conditions of a 240MHz sampling rate and a noise phase walk coefficient of 0.3, a local signal with a length of 65536 points and a constant frequency of 10MHz is constructed. To verify the effectiveness of the algorithm in scenarios involving complex high-order target motion, this embodiment constructs a physical motion model of the target in the simulation to generate the echo signal for testing. Specifically, the initial motion state of the target is used as a baseline for calculating subsequent velocities. In this physical model, the initial velocity is set to 20m / s, the acceleration to -88.2m / s², and the jerk to -88.2m / s³. Considering that in practical coherent lidar systems, the noise of the echo signal can usually be equivalently modeled as additive white Gaussian noise, and to verify the effectiveness of the algorithm of this invention under extremely low signal-to-noise ratio and weak signal conditions, additive white Gaussian noise was superimposed on the signal in the simulation to construct an echo target signal with a variable speed motion and a signal-to-noise ratio of -26dB, simulating the generation of severe Doppler migration and a weak echo signal. Figure 2 The image shows the Doppler frequency shift of the target, indicating that the Doppler frequency shift is constantly changing. Then, eight segments of non-coherent accumulation were performed on the signal, and the accumulation results with and without Doppler migration compensation were compared. The results are as follows: Figure 3 As shown, the amplitude spectrum obtained by direct accumulation does not form the expected narrowband peak; instead, the spectrum broadens, resulting in severe energy defocusing. The signal-to-noise ratio (SNR) calculated by direct accumulation is only 2.86 dB, and the signal peak is flat, making it difficult to detect stably in complex noise backgrounds. However, for the accumulated spectrum after Doppler migration compensation, the energy is focused at the theoretically defined Doppler frequency, forming an extremely clear, sharp, and prominent energy peak. The SNR is improved to 12.56 dB, an improvement of 9.7 dB compared to the spectrum obtained by direct accumulation. This demonstrates that the method can effectively estimate and compensate for Doppler migration, achieving energy focusing.

[0049] This invention embodiment conducted experiments on 20 sets of data, with each set undergoing 8 segments of non-coherent accumulation. The cumulative signal-to-noise ratio results for the 20 sets of data are as follows: Figure 4 As shown, the compensated cumulative signal-to-noise ratio is significantly higher than that of the direct cumulative signal-to-noise ratio, with an average difference of 8.67 dB, which confirms the effectiveness of the method of the present invention.

[0050] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A Doppler migration compensation method based on phase compensation and front-end pulse feedback, characterized in that, Includes the following steps: Step S1: Perform Hilbert construction on the local signal to form a complex signal, and take its conjugate to obtain the local complex conjugate signal; Step S2: Mix the local complex conjugate signal with the received echo signal to obtain a mixed signal with pulse compression and Doppler frequency shift extraction. Step S3: Calculate the peak frequency of the spectrum of the mixed signal, i.e., the Doppler frequency shift, and save the Doppler frequency shift; Step S4: Determine whether the signal being processed is the first segment of the current accumulation block. If yes, perform non-coherent accumulation on the spectrum of the current signal directly. If no, proceed to steps S5 to S8. Step S5: Extract the Doppler frequency shift saved from the previous mixing signal to construct the phase compensation function; Step S6: Apply the phase compensation function to perform phase compensation on the mixing signal, and calculate the peak frequency of the spectrum of the phase-compensated signal, which is the number of Doppler migration units of the target. Step S7: Cyclicly shift the historical cumulative spectrum according to the number of Doppler migration units of the target to compensate for Doppler migration; Step S8: Finally, the compensated historical cumulative spectrum is non-coherently accumulated with the current segment mixing signal spectrum.

2. The Doppler migration compensation method based on phase compensation and front-end pulse feedback as described in claim 1, characterized in that, The local signal in step S1 is represented as ,in It is the local signal amplitude. It's phase noise. The local frequency is used to construct a complex signal by performing a Hilbert transform on the local signal, and then taking its conjugate to obtain the local complex conjugate signal. .

3. The Doppler migration compensation method based on phase compensation and front-end pulse feedback as described in claim 1, characterized in that, In step S2, the local complex conjugate signal is mixed with the received echo signal to obtain a pulse-compressed and Doppler-shift-extracted mixed signal, specifically including: The local complexized conjugate signal is represented as: ,in It is the local signal amplitude. It is a local frequency. It is the phase noise of the signal, and the echo signal is represented as... ,in It is the echo amplitude. It's a Doppler shift. It is signal noise. Multiplying the local complex conjugate signal with the echo signal and mixing them, the resulting mixed signal is: ,in It is the amplitude of the mixing signal. It is the noise cross term in the mixing process, that is, the noise of the mixing signal.

4. The Doppler migration compensation method based on phase compensation and front-end pulse feedback as described in claim 1, characterized in that, The Doppler frequency shift refers to the reference Doppler frequency shift value used for compensation in the next data segment. .

5. The Doppler migration compensation method based on phase compensation and front-end pulse feedback as described in claim 1, characterized in that, Step S5: Extract the Doppler frequency shift saved from the previous mixing signal to construct the phase compensation function, the mathematical expression of which is: ,in, The phase compensation function is... It is the imaginary unit.

6. The Doppler migration compensation method based on phase compensation and front-end pulse feedback as described in claim 1, characterized in that, In step S6, applying the phase compensation function to perform phase compensation on the mixing signal and calculating the number of Doppler migration units of the target specifically includes: applying the phase compensation function... Multiply the current segment's mixing signal by a complex number to obtain the compensated mixing signal; Perform a Fourier transform on the compensated mixed signal to obtain the compensated amplitude spectrum, and search for the peak index positions of this compensated amplitude spectrum starting from 0. ; The number of Doppler migratory units is calculated based on the peak index position. .

7. The Doppler migration compensation method based on phase compensation and front-end pulse feedback as described in claim 6, characterized in that, The number of Doppler migratory units is calculated based on the peak index position. The calculation rule is as follows: when hour, ;when hour, ,in This represents the number of sampling points for a single signal segment.

8. The Doppler migration compensation method based on phase compensation and front-end pulse feedback as described in claim 1, characterized in that, In step S7, the historical cumulative spectrum is cyclically shifted according to the number of Doppler migrating units of the target. The cyclic shifting method is to move towards the latest target Doppler frequency point based on the estimated number of target Doppler migrating units.

9. The Doppler migration compensation method based on phase compensation and front-end pulse feedback as described in claim 8, characterized in that, The specific circular shift method is as follows: If the value is greater than 0, then shift to the right in a circular fashion. Bit, If the value is less than 0, then shift to the left in a circular shift. Bit.

Images