Invention content
Object of the present invention is to provide an FMCW radar super-resolution range imaging method and apparatus to solve the problems raised in the above background technology.
In order to solve the above technical problems, the present invention provides the following technical solution: an FMCW radar super-resolution range imaging method, comprising the following steps:
Step 1: Spectral peak rough estimation to obtain the spectral peak position of the sampling sequence of the beating signal; The differential beat signal is the signal obtained after mixing the FMCW radar target echo and the transmitted signal, and the sampling sequence of the bad beat signal is represented by s, N is the number of samples, representing the complex matrix of N×1 dimension,![]()
![]()
and the mathematical symbol representing the matrix dimension and type;
Step 2: Refine the frequency of the interval of interest, refine the frequency within a certain range around the spectral peak position obtained by the spectral peak coarsely estimated, and the frequency point of the refined interval of interest is L is![]()
the number of frequency points of the refined interval of interest, and form a popular matrix![]()
of the frequency domain of the interval of interest as shown in the following formula:
A＝exp(j2πfxhTt)T (1)
In formula (1), t is the sampling time sequence, which is a vector of 1×N, j is a complex symbol, and T represents the matrix transpose operator;
Step 3: Initialize the spectrum of the interval of interest, use the popular matrix A in step 2 as the initial value of the filter weight matrix, and obtain the refined spectrum initialization result![]()
as follows:
y0＝AHs (2)
In equation (2), H represents the matrix conjugate transpose operator;
The interval frequency of interest is the frequency of interest. After the rough estimation of the peak, you can know what frequency/distance the target is in (distance and frequency are one-to-one correspondence), and these potential targeted frequency/distance ranges need to be focused on, so fine imaging is refined for these frequencies. This step also discards the frequency of no interest, that is, the frequency/distance without the target, saving computing power.
Step 4: Filter weight optimization update, update the filter weight based on the MMSE criterion, and the cost function of the MMSE criterion is
J＝E{|| y-WHs|| 2} (3)
In equation (3), E{} represents the expectation operator, y is the spectrum of the beatbeat signal, and equation (3) is used to derive W and equal W to zero, and obtain the optimal weight vector, as shown in the following equation:
W＝(APAH+Rv)AP (4)
In Equation (4), P=[yyH]⊙I L×L, ⊙ represents the Hadamard product, ILxL represents the identity matrix of L×L, P is obtained from the spectral y obtained by initialization or the previous iteration, and Rv is the noise covariance matrix;
Step 5: Interval of Interest spectrum update, update the spectrum estimation results with the updated filter weights, as shown in the following equation:
y＝WHs (5)
Step 6: Iteration end condition judgment, repeat iteration steps 6~7 until the number of iterations or end conditions are met;
Each iteration of the imaging effect will become a little better, set the end of the iteration condition is a compromise to consider the effect and real-time, if you feel that the iteration is almost the same number of iterations, you can also set a fixed number of iterations, you can also judge whether to terminate the loop iteration according to the improvement after the iteration, if this iteration and the last time is only a little better, then it is considered that the performance has basically converged, and there is no need to iterate the cycle.
Step 7: Frequency distance conversion, according to the correspondence between frequency and distance, the FMCW radar range super-resolution result is obtained, as shown in the following formula:
![]()
In Equation (6), c is the electromagnetic wave propagation speed, fr is the modulation frequency of FMCW radar, and f1, f2,...,f L are different refinement frequencies.
FMCW radar frequency and distance are one-to-one correspondence, the correspondence is that the difference signal frequency f corresponds to the target distance after![]()
the previous iteration, and the spectrum is obtained, which is the spectral amplitude y of different f (refinement frequency points f1, f2,...,f L), which is equivalent to the target size at the distance![]()
.
Further, a fast Fourier transform FFT was used in step 1 for peak coarse estimation.
The present invention provides an FMCW radar super-resolution range imaging device, including a spectral peak coarse estimation module, an interest interval frequency refinement module, an interest interval spectrum initialization module, a filter weight optimization update module, an interest interval spectrum update module, an iteration end condition decision module, and a frequency distance conversion module;
The Spectral Peak Coarse Estimation Module is used to obtain the spectral peak position of the sampling sequence of the beat signal; The differential beat signal is the signal obtained after mixing the FMCW radar target echo and the transmitted signal, the sampling sequence of the bad beat signal is represented by s, N is the number of samples,![]()
![]()
representing the complex matrix of N×1 dimension, and the mathematical symbol representing the matrix dimension and type;
The interval of interest frequency refinement module is used to refine the frequency within a certain range around the spectral peak position obtained by the spectral peak coarse estimation, and the refined frequency point of the interest interval frequency point is L is the number of frequency points of the refined interval of interest, and the popular matrix![]()
of the frequency domain of the interval of interest is![]()
formed as shown in the following formula:
A＝exp(j2πfxhTt)T (1)
In formula (1), t is the sampling time sequence, which is a vector of 1×N, j is a complex symbol, and T represents the matrix transpose operator;
In the spectrum initialization module of the interval of interest, the popular matrix is used as the initial value of the filter weight matrix, and the refinement spectrum initialization result![]()
is obtained as follows:
y0＝AHs (2)
In equation (2), H represents the matrix conjugate transpose operator;
The interval frequency of interest is the frequency of interest. After the rough estimation of the peak, you can know what frequency/distance the target is in (distance and frequency are one-to-one correspondence), and these potential targeted frequency/distance ranges need to be focused on, so fine imaging is refined for these frequencies. This step also discards the frequency of no interest, that is, the frequency/distance without the target, saving computing power.
The filter weight optimization update module updates the filter weight based on the MMSE criterion, and the cost function of the MMSE criterion is
J＝E{|| y-WHs|| 2} (3)
In equation (3), E{} represents the expectation operator, y is the spectrum of the beatbeat signal, and equation (3) is used to derive W and equal W to zero, and obtain the optimal weight vector, as shown in the following equation:
W＝(APAH+Rv)AP (4)
In equation (4), P=[yyH]⊙I L×L, ⊙ represents the Hadamard product, I L×L represents the identity matrix of L×L, P is obtained from the spectrum y obtained by initialization or the previous iteration, and Rv is the noise covariance matrix;
The Interval of Interest Spectrum Update module is used to update the spectrum estimation results with updated filter weights, as shown in the following equation:
y＝WHs (5)
The iteration end condition judgment module determines whether the number of iterations or the iteration end condition is met, and stops the iteration, otherwise continue to iterate to update the filter weight and spectrum estimate;
The frequency distance conversion module is used to obtain the FMCW radar range super-resolution result according to the correspondence between frequency and distance, as shown in the following equation:
![]()
In Equation (6), c is the electromagnetic wave propagation speed, fr is the modulation frequency of FMCW radar, and f1, f2,...,f L are different refinement frequency points.
FMCW radar frequency and distance are one-to-one correspondence, the correspondence is that the difference signal frequency f corresponds to the target distance after![]()
the previous iteration, and the spectrum is obtained, which is the spectral amplitude y y of different f (refinement frequency points f1, f2,...,f L), which is equivalent to the target size at the distance![]()
.
The present invention provides a computer-readable access medium, a computer-readable access medium is stored on a computer program, and the computer program is executed by the processor to implement the above FMCW radar super-resolution range imaging method.
Further, the Spectral Peak Coarse Estimation Module uses fast Fourier transform FFT for peak coarse estimation.
Compared with the prior art, the beneficial effects achieved by the present invention are:
1. By constructing a frequency domain manifold matrix, the invention applies the iterative adaptive super-resolution algorithm to the frequency estimation of FMCW radar difference frequency signal, combined with the refined frequency interval, can effectively suppress the estimation value of the untargeted distance unit/frequency unit, reduce the flooding effect of the strong target distance sidelobe on the nearby small target, and obtain a higher resolution range image.
2. The present invention determines the coarse position of the target through spectral coarse estimation, and only performs frequency refinement near the coarse position, which greatly reduces the size of the frequency domain of the interval of interest, reduces the matrix dimension of the inverse operation and multiplication and addition operation, and greatly reduces the amount of operation.
The accompanying drawings further describe the conception, specific structure and technical effects of the present invention to fully understand the object, characteristics and effects of the present invention.
Specific embodiment
The present invention is further elaborated below in conjunction with specific embodiments. It should be understood that these embodiments are intended only to illustrate the present invention and are not intended to limit the scope of the present invention. Further, it should be understood that after reading the content of the present invention, those skilled in the art may make various modifications or modifications to the present invention, and these equivalent forms also fall within the scope of the claims appended to the present application.
In the drawings, structurally identical components are indicated by identical numerical designators, and components with similar structures or functions everywhere are indicated by similar numerical designators. The size and thickness of each component shown in the drawings are shown arbitrarily, and the present invention does not limit the size and thickness of each component. In order to make the illustration clearer, the thickness of the part is appropriately exaggerated in some places in the drawings.
As shown in FIG. 1 is a step-by-step flow chart of an FMCW radar super-resolution range imaging method provided by the present invention, embodiments provided by the present invention are carried out in accordance with the step-by-step flow shown in FIG. 1.
The present invention provides a process of super-resolution processing of FMCW radar bad beat signal, FMCW radar transmits chirp signal, frequency range of 77.0GHz-77.3GHz, frequency modulation fr is 10MHz/us, bandwidth 300MHz, single transmission signal pulse width 30us, radar receives signal and transmit signal mixed to obtain a bad beat signal, the sampling rate of the bad beat signal is 30Msps, and the sampling sequence of the bad beat signal is ![]()
y is the actual spectrum of the beater signal, as follows:
![]()
In Equation (7), y(f l) is the amplitude of the frequency component fl, and L0 is the number of all frequency components that make up the beat signal.
The popular matrix defining the spectrum of the interval of interest is as follows:
![]()
ts is the sampling period;
The sampling sequence of the beat signal is represented by s as follows:
s＝Ay+v (8)
In Equation (8), v is the noise sequence of N×1.
The steps of FMCW radar super-resolution range imaging method are as follows:
Step 1: Spectral peak rough estimation, use the FPGA built-in mip core to realize the FFT of the above bad beat signal sampling sequence s, obtain the rough spectrum![]()
of the bad beat signal, use CFAR to detect and search the spectral peak position of the bad beat signal, f1, f2,...,f K,K is the number of spectral peaks;
Step 2: The frequency of the interval of interest is refined, and the frequency within a certain range around the spectral peak position obtained by the spectral peak coarse estimation is refined 10 times, and the frequency point of the refined interval of interest is L is![]()
the frequency frequency point of the refined interval of interest.
fxh＝[f1-3Δf：Δf/10：f1+3Δf，f2-3Δf：Δf/10：f2+3Δf，...，fK-3Δf：Δf/10：fK+3Δf] (9)
In Equation (9), Δf is the discrete frequency interval in the coarse spectrum, Δf = 1/T r, and Tr is the pulse width of the single transmitted signal.
In the present embodiment, when the frequency around a certain range of the spectral peak position obtained by the rough estimation of the spectral peak is refined, the frequency is 10 times the frequency of 3 Δf on the left and right of the frequency corresponding to the rough estimate, and the frequency of the frequency corresponding to the spectral peak can also be refined by 1 Δf, 2 Δf or 4 Δf on the left and right of the frequency corresponding to the peak.
The frequency points in fxh are deduplicated to remove overlapping frequency points. The popular matrix![]()
that forms the frequency domain of the interval of interest is shown in the following equation:
A＝exp(j2πfxhTt)T (1)
In formula (1), t is the sampling time sequence, which is a vector of 1×N, j is a complex symbol, and T represents the matrix transpose operator;
Step 3: Initialize the spectrum of the interval of interest, use the popular matrix A in step 2 as the initial value of the filter weight matrix, and obtain the refined spectrum initialization result![]()
as follows:
y0＝AHs (2)
In equation (2), H represents the matrix conjugate transpose operator;
Step 4: Filter weight optimization update, using the initialization result and noise covariance matrix, update the filter weight based on the MMSE criterion, and the cost function of the MMSE criterion is used
J＝E{|| y-WHs|| 2} (3)
In equation (3), E{} represents the expectation operator, y is the spectrum of the beatbeat signal, and equation (3) is used to derive W and equal W to zero, and obtain the optimal weight vector, as shown in the following equation:
W＝(E{ssH})-1E{syH} (9)
Equation (8) s=Ay+v is brought into equation (9) and obtained after simplification
W＝(APAH+Rv)AP (4)
In equation (4), P=[yyH]⊙I L×L, ⊙ represents the Hadamard product, I L×L represents the identity matrix of L×L, P is obtained from the spectrum y obtained by initialization or the previous iteration, R v is the noise covariance matrix, Rv = var· IN×N；
Step 5: Interval of Interest spectrum update, update the spectrum estimation results with the updated filter weights, as shown in the following equation:
y＝WHs (5)
Step 6: Iteration end condition judgment, repeat iteration steps 6~7 until the number of iterations or end conditions are met;
Step 7: Frequency distance conversion, according to the correspondence between frequency and distance, the FMCW radar range super-resolution result is obtained, as shown in the following formula:
![]()
In Equation (6), c is the electromagnetic wave propagation speed, fr is the modulation frequency of FMCW radar, and f1, f2,...,f L are different refinement frequencies.
FMCW radar frequency and distance are one-to-one correspondence, the correspondence is that the difference signal frequency f corresponds to the target distance after![]()
the previous iteration, and the spectrum is obtained, which is the spectral amplitude y y of different f (refinement frequency points f1, f2,...,f L), which is equivalent to the target size at the distance![]()
.
Example 1
As shown in Figure 2, using the above FMCW radar super-resolution range imaging method, the point target distance image results under the conditions of distance 30m, scattering point amplitude 1, SNR=10dB are given.
Example 2
As shown in Figure 3, using the above FMCW radar super-resolution range imaging method, the distance image results of two adjacent targets (distances of 29.7m and 30m, scattering point amplitude of 0.1 and 1, SNR=10dB) are given.
Example three
As shown in Figure 4, using the above-mentioned FMCW radar super-resolution range imaging method, the approximate continuous target (distance 30~45m, 1 strong scattering point per interval of 0.6m, scattering point amplitude is 1, SNR=10dB) within a distance.
Figure 2~4, the line corresponding to the FFT result is the output result of step 1, the line corresponding to the initialization result is the output result of step 3, the line corresponding to the iterative adaptive result is the result after the end of step 6 iteration, the narrower and sharper the spike at the corresponding target distance (abscissa), the better the resolution, when the two targets are close together, two independent peaks can be separated to indicate good resolution, so it can be concluded that the iterative adaptive distance imaging method used in the present application is used to estimate the frequency of FMCW radar difference frequency signal. Higher resolution distance images can be obtained.
In Examples 1 to 3, the frequency points of the full frequency domain before refinement are 900, and 9000 after refinement according to 10 times, and the frequency refinement frequency points of the region of interest of Examples 1, 2 and 3 after rough estimation of the spectrum are 60, 60, and 360 respectively. It can be seen that after rough spectral estimation, the matrix dimension participating in the operation in the iterative adaptation process can be greatly reduced, the computational burden can be reduced, and the calculation speed of the algorithm can be improved.
Example IV
The present invention provides an FMCW radar super-resolution range imaging device, including a spectral peak coarse estimation module, an interest interval frequency refinement module, an interest interval spectrum initialization module, a filter weight optimization update module, an interest interval spectrum update module, an iteration end condition decision module, and a frequency distance conversion module;
Spectral peak rough estimation module, based on the FFT IP core and CFAR module that comes with the FPGA, realizes FFT of the bad beat signal, obtains the rough spectrum of the bad beat signal, and uses CFAR detection to search the spectral peak position of the bad beat signal, the sampling sequence of the bad beat signal is represented by s, N is the number of samples, representing the complex matrix of N×1 dimension,![]()
![]()
and the mathematical symbol representing the matrix dimension and type;
The interval of interest frequency refinement module is used to refine the frequency within a certain range around the spectral peak position obtained by the spectral peak coarse estimation, and the refined frequency point of the interest interval frequency point is L is the number of frequency points of the refined interval of interest, and the popular matrix![]()
of the frequency domain of the interval of interest is![]()
formed as shown in the following formula:
A＝exp(j2πfxhTt)T (1)
In formula (1), t is the sampling time sequence, which is a vector of 1×N, j is a complex symbol, and T represents the matrix transpose operator;
In the spectrum initialization module of the interval of interest, the popular matrix is used as the initial value of the filter weight matrix, and the refinement spectrum initialization result![]()
is obtained as follows:
y0＝AHs (2)
In equation (2), H represents the matrix conjugate transpose operator;
The interval frequency of interest is the frequency of interest. After the rough estimation of the peak, you can know what frequency/distance the target is in (distance and frequency are one-to-one correspondence), and these potential targeted frequency/distance ranges need to be focused on, so fine imaging is refined for these frequencies. This step also discards the frequency of no interest, that is, the frequency/distance without the target, saving computing power.
The filter weight optimization update module updates the filter weight based on the MMSE criterion, and the cost function of the MMSE criterion is
J＝E{|| y-WHs|| 2} (3)
In equation (3), E{} represents the expectation operator, y is the spectrum of the beatbeat signal, and equation (3) is used to derive W and equal W to zero, and obtain the optimal weight vector, as shown in the following equation:
W＝(APAH+Rv)AP (4)
In equation (4), P = [yyH]⊙I L×L, ⊙ represents the Hadamard product, I L×L represents the identity matrix of L×L, P is obtained from the spectrum y obtained by initialization or the previous iteration, and Rv is the noise covariance matrix;
The Interval of Interest Spectrum Update module is used to update the spectrum estimation results with updated filter weights, as shown in the following equation:
y＝WHs (5)
The iteration end condition judgment module determines whether the number of iterations or the iteration end condition is met, and stops the iteration, otherwise continue to iterate to update the filter weight and spectrum estimate;
The frequency distance conversion module is used to obtain the FMCW radar range super-resolution result according to the correspondence between frequency and distance, as shown in the following equation:
![]()
In Equation (6), c is the electromagnetic wave propagation speed, fr is the modulation frequency of FMCW radar, and f1, f2,...,f L are different refinement frequencies.
FMCW radar frequency and distance are one-to-one correspondence, the correspondence is that the difference signal frequency f corresponds to the target distance after![]()
the previous iteration, and the spectrum is obtained, which is the spectral amplitude y y of different f (refinement frequency points f1, f2,...,f L), which is equivalent to the target size at the distance![]()
.
The present invention also provides a computer-readable access medium, a computer-readable access medium is stored on a computer program, and the computer program is executed by the processor to implement the FMCW radar super-resolution range imaging method shown in FIG. 1.
The above embodiments only express several embodiments of the present invention, and their description is more specific and detailed, but it cannot be understood as a limitation on the scope of the patent of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the idea of the present invention, a number of deformations and improvements may also be made, which fall within the scope of protection of the present invention. Therefore, the scope of protection of the invention patent shall be subject to the attached claims.
400-4929-0909
contact@catarc.com.cn
