CN109471095B - FMCW radar distance estimation method based on fast iterative interpolation - Google Patents
FMCW radar distance estimation method based on fast iterative interpolation Download PDFInfo
- Publication number
- CN109471095B CN109471095B CN201811310159.8A CN201811310159A CN109471095B CN 109471095 B CN109471095 B CN 109471095B CN 201811310159 A CN201811310159 A CN 201811310159A CN 109471095 B CN109471095 B CN 109471095B
- Authority
- CN
- China
- Prior art keywords
- target
- spectral
- deviation
- frequency
- unit
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/06—Systems determining position data of a target
- G01S13/08—Systems for measuring distance only
- G01S13/32—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated
Abstract
The invention relates to an FMCW radar distance estimation algorithm based on fast iterative interpolation, which aims at the problem that the ranging precision of an on-vehicle anti-collision radar in a multi-target environment is easily interfered by frequency spectrums. The algorithm provided by the simulation verification has higher estimation accuracy in a multi-target environment and under the condition of low signal to noise ratio. And finally, the algorithm is realized by DSP, and the algorithm is proved to meet the requirement of real-time property and is easy to realize in engineering.
Description
Technical Field
The invention relates to an FMCW radar distance estimation algorithm, in particular to an FMCW radar distance estimation algorithm based on fast iterative interpolation, belongs to the technical field of millimeter wave radars and the field of signal processing, and particularly relates to interference among multiple targets.
Background
Frequency Modulated Continuous Wave (FMCW) radar has good speed and distance measuring performance, and thus is widely applied to the aspects of unmanned driving, medical monitoring, channel estimation and the like. However, the conventional FMCW radar has the problems of fence effect and spectrum leakage, which causes large estimation deviation of the distance. For more accurate estimation of the target distance, the current common methods are: a sampling sequence direct zero padding method, a Zoom Fast Fourier Transform (ZFFT) method based on complex modulation, a Fourier Transform-Discrete-time Fourier Transform (FFT-DTFT) combined correction method, a Chirp-Z Transform, and the like. However, the above four methods have the disadvantages of large calculation amount, large estimation error and the like.
In recent years, some frequency estimation methods based on maximum likelihood estimation have been proposed, which can also be used for distance estimation. The document (an improved ZFT algorithm for improving the LFMCW radar ranging accuracy, 2015, (2): 1-4.) provides an iterative interpolation algorithm under a single target signal, and the deviation is gradually corrected in the iterative process by using an interpolation result; the literature (Accurate Frequency Estimators,2007,24 (1): 123-125.) proposes a Frequency deviation correction method based on Discrete Fourier Transform (DFT) samples, which does not need iteration and directly estimates the deviation by using the adjacent spectral peak with the target; on the basis, a deviation correction factor is added in a document (A Method For Fine Resolution timing From Three DFT Samples, IEEE Signal Processing Letters,2011,18 (6): 351-354), so that the precision is improved, but the problem of non-uniform Estimation error exists; the document (frequency iterative interpolation estimation algorithm based on the refined frequency spectrum, electronic and informatics newspaper, 2017,39 (9): 2141-2147.) proposes a method for interpolating signals by using the refined frequency spectrum, thereby solving the problem of non-uniform estimation errors. The method proposed by the above document has a good effect on the deviation estimation of the single-frequency signal, but when the method is applied to a multi-target environment, the deviation of the distance estimation is increased due to inter-spectrum interference. In order to solve the problem, the literature (DFT Interpolation Algorithm for Kaiser-Bessel and Dolph-Chebyshev Windows, IEEE Transactions on Instrumentation & Measurement,2011,60 (3): 784-790) proposes a method for compensating estimation deviation based on high-order polynomial Interpolation, and the method uses high-order polynomial to fit the frequency spectrum of the signal to obtain the accurate frequency value of each target, but the Algorithm has large calculation amount and is not beneficial to hardware implementation. The document (Fine resolution frequency estimation from the same DFT samples: case of windowed data, signal Processing,2015,114 (1): 245-250.) proposes a method for suppressing spectral interference by windowing, and a window function is used for suppressing interference of sidelobes, but the method proposed by the document has the problem that spectral broadening influences estimation performance. In the field of DOA estimation, the literature (Fast iterative polarized beamforming for acquisition single-snapshot DOA estimation, IEEE Geoscience and Remote Sensing Letters,2017,14 (4): 574-578) proposes a method for iteratively and gradually correcting angular deviation, which can avoid the problem of spectrum broadening and solve the interference between multiple target angles, but the method applied to an FMCW radar system has the following problems: in the course of rough estimation, the number of targets is used as prior information, and the original data is required to be corrected when each target spectral peak is estimated, thereby increasing the complexity of the algorithm.
The invention provides an improved FMCW radar distance estimation algorithm based on fast iterative interpolation, which is inspired by solving the spectrum interference under the condition of multiple targets by the iterative correction idea, roughly estimates the position of the target by a constant false alarm detection method to solve the problem that the number of the targets is used as prior information, and replaces the process of repeatedly correcting the original data by using a correction deviation factor in the iterative process. The invention is used for solving the problem of inter-spectrum interference in a multi-target environment, in particular to a method which can eliminate the inter-spectrum interference of multiple targets and obtain more accurate target reflection signal frequency and target distance. The processing process is simplified, and the hardware implementation is facilitated. Theoretical simulation and actual measurement data verify the feasibility of the algorithm.
Disclosure of Invention
In view of the above prior art, the technical problem to be solved by the present invention is to provide an FMCW radar distance estimation algorithm based on fast iterative interpolation, which can overcome the influence of interference between frequency spectrums in a multi-target environment on frequency estimation, so as to estimate the frequency deviation of a target by using the reflected signal of the target itself.
In order to solve the technical problem, the FMCW radar distance estimation algorithm based on fast iterative interpolation comprises the following steps:
step 1: carrying out Fourier transformation on the intermediate frequency signals after AD sampling, and then carrying out constant false alarm detection to estimate the number P of targets and the frequency spectrum unit m where the targets are located i ,i＝1,2...P；
Step 2: calculating the spectral unit m of the ith target i Corresponding spectral cell deviation
i =1,2 \ 8230p, step 2 comprising the steps of:
step 2.1: setting a target iteration number Q, and enabling the current iteration number Q =0; deviation estimate for ith target after single iteration
And amplitude estimator
Carry out initialization to
i＝1,2…P；
Step 2.2: estimating the reflected signal strength of the p-th target
p＝1,2...P；
Step 2.3: by using
Calculating the frequency spectrum interference generated by the p target to the i target
Step 2.4: after calculating and removing interference
Spectral peaks of spectral units
Step 2.5: using spectral peaks
Calculating a deviation correction factor
Deviation from the spectral unit
Step 2.6: q ← Q +1, judging whether Q is smaller than Q, if Q is smaller than or equal to Q, repeating the steps 2.2-2.5; if Q is greater than Q, executing step 3;
and step 3: using spectral units m i And spectral unit deviation
Calculate the ithFrequency f of target reflected signal i And the distance value R of the target i 。
The invention also includes:
1. in step 1, the intermediate frequency signal after AD sampling is subjected to fourier transform expression as follows:
wherein s is b (N) represents an intermediate frequency signal after AD sampling, N represents the number of points of Fourier transform, S b (k) And expressing the amplitude of the kth unit in the frequency domain, and estimating that the expression of the target number P satisfies the following conditions:
P＝Count{S b (k)＞CFAR(S b (k))}
wherein CFAR (S) b (k) Represents a pair S b (k) The result of the constant false alarm detection, count represents the S pair b (k) Greater than CFAR (S) b (k) The number of the amplitude values is counted, and the counted result is the number of the targets;
the spectral unit m where the p-th target is located p The expression is as follows:
2. the number of iterations Q in step 2.1 is set to 2.
3. Intensity of reflected signal of p-th target in step 2.2
The estimated value expression of (c) satisfies:
wherein
Is the estimated value of the intensity of the reflected signal of the first target, and the initial value is 0，Δm pl Indicating the frequency spectrum unit m corresponding to the p-th target p Spectral unit m corresponding to the l-th target l At a distance in the frequency domain, the expression satisfies:
Δm pl ＝Δm p -Δm l
Δδ pl indicating the deviation of the spectrum unit corresponding to the p-th target
Deviation of spectral unit corresponding to the l-th target
The distance in the frequency domain is expressed as follows:
Δδ pl ＝Δδ p -Δδ l 。
4. step 2.3, the p-th target generates spectrum interference to the i-th target
The expression satisfies:
wherein Δ m ip Indicating the frequency spectrum unit m corresponding to the ith target i M of the spectrum unit corresponding to the p-th target p At a distance in the frequency domain, the expression satisfies:
Δm ip ＝Δm i -Δm p
Δδ ip indicating the deviation of the spectrum unit corresponding to the ith target
Deviation of spectral unit corresponding to p-th target
At a distance in the frequency domain, the expression satisfies:
Δδ ip ＝Δδ i -Δδ p 。
5. after the interference is removed in step 2.4
Spectral peaks of spectral units
The expression satisfies:
wherein S i,±0.5 Representing direct Fourier transform of an intermediate frequency signal
Spectral peak of spectral unit, S i,±0.5 The device consists of two parts: the ith target reflected signal is
The spectrum peak and the p target obtained by the Fourier transform of the spectrum unit generate spectrum interference on the i target, and the expression satisfies the following conditions:
6. deviation correction factor in step 2.5
Deviation from the spectral unit
The expressions respectively satisfy:
7. frequency f of ith target reflection signal in step 3 i And the distance value R of the target i The expressions respectively satisfy:
wherein f is s Is the AD sampling frequency, c is the speed of light, B is the bandwidth of the modulation, T is the period of the modulated triangular wave.
The invention has the beneficial effects that:
the method firstly estimates the number of the targets and the rough position of the targets by utilizing fast Fourier transform and combining a constant false alarm detection algorithm, then carries out non-uniform iterative interpolation on the frequency spectrum of the targets, and eliminates interference between the frequency spectrums in the process of each iteration, thereby obtaining more accurate distance.
Drawings
FIG. 1 is a block diagram of an FMCW radar system;
FIG. 2 is the distance root mean square error of the target 1 at different signal-to-noise ratios;
FIG. 3 is the distance root mean square error of the target 2 at different signal-to-noise ratios;
FIG. 4 is the effect of target 2 sampling point on distance root mean square error;
FIG. 5 is a plot of distance root mean square error for different algorithms under multiple objectives;
FIG. 6 is the result of a local refinement of the target 3;
FIG. 7 is the result of a local refinement of the target 8;
FIG. 8 is a real environment test chart;
FIG. 9 is an FMCW radar system;
fig. 10 is the result of FMCW radar output versus the target true distance.
Detailed Description
Let S be for convenience of description i ＝S b (m i )，S b (m i ) Represents the m-th in the frequency domain i Spectral peaks of the individual spectral units, then
The spectral peaks of individual spectral units can be expressed as:
the specific implementation mode for realizing the aim of the invention is as follows:
step 1: performing Fast Fourier Transformation (FFT) on the intermediate frequency signal after AD sampling, and then performing Constant False Alarm detection (CFAR) to estimate the number P of targets and the spectrum unit m where the targets are located i ,i＝1,2...P。
Fourier transform expression is as follows
Wherein s is b (N) represents the intermediate frequency signal after AD sampling, N represents the number of points of FFT conversion, S b (k) Representing the amplitude of the k-th element in the frequency domain. By pair S b (k) The power level of the background noise is estimated by CFAR processing, and then S is calculated b (k) Comparing with the background noise power level to estimate the target number P, the expression is as follows
P＝Count{S b (k)＞CFAR(S b (k))}
Wherein CFAR (S) b (k) Represents a pair S b (k) The result of the constant false alarm detection, i.e. the background noise power, count represents the pair S b (k) Greater than CFAR (S) b (k) The number of the spectrum peaks is counted, and the counted result is the number of the targets. In addition, it is necessary to extract the spectrum unit where the target is located and use the extracted spectrum unit as the estimated spectrum unit deviation
And (4) a foundation.
The spectral unit m in which the p-th target is located p The expression is as follows
And 2, step: calculating the spectral unit m of the ith target i Corresponding deviation of spectral unit
i＝1,2…P。
Step 2.1: setting a target iteration number Q, and enabling the current iteration number Q =0; deviation estimate for ith target after single iteration
And the intensity of the reflected signal of the ith target
Carry out initialization to
i＝1,2…P。
The iteration times Q is set to be 2, experiments show that when Q is 2, high distance precision can be achieved, and the distance estimation precision is slightly influenced when multiple iterations are carried out. And when the iteration is performed for multiple times, only the calculated amount is increased, and the influence on the distance estimation precision is small.
Step 2.2: estimating the reflected signal strength of the p-th target
p＝1,2...P。
Directly using intermediate frequency signals at
The spectrum peak obtained by FFT conversion of the spectrum unit comprises three parts: the p-th target reflection signal is processed by FFT to obtain the amplitude, and other target reflection signals are processed by FFT
Leakage of spectral cells and noise-to-amplitude interference, where the interference of noise has a less negligible effect on the spectral peaks. The first target is
The leakage of the spectrum unit can be reflected at the target through the target
The position of the frequency spectrum unit is obtained by FFT, and the expression is as follows
Wherein
Is an estimate of the reflected signal strength of the first target, with an initial value of 0,s bl And (n) represents an intermediate frequency signal reflected by the l target. Due to s bl (n) is not available, therefore
The process of the true value is also continuously approached by iterating the successive updates. Δ m pl Indicating the frequency spectrum unit m corresponding to the p-th target p M of the spectrum unit corresponding to the l-th target l The distance in the frequency domain is expressed as follows
Δm pl ＝Δm p -Δm l
Δδ pl Indicating the deviation of the spectrum unit corresponding to the p-th target
Deviation of spectral unit corresponding to the l-th target
The distance in the frequency domain is expressed as follows
Δδ pl ＝Δδ p -Δδ l
Thereby obtaining the reflected signal intensity of the p-th target
The expression is as follows:
step 2.3: by using
Calculating the frequency spectrum interference generated by the p target to the i target
The p target generates spectrum interference to the i target
Can pass through the p-th target intermediate frequency signal s bp (n) is in
The spectrum unit is obtained by FFT conversion, but s bp (n) is not directly available and therefore needs to be utilized
Obtaining
The expression is as follows
Wherein Δ m ip Indicating the frequency spectrum unit m corresponding to the ith target i M of the spectrum unit corresponding to the p-th target p The distance in the frequency domain is expressed as follows
Δm ip ＝Δm i -Δm p
Δδ ip Indicating the deviation of the spectrum unit corresponding to the ith target
Deviation of spectral unit corresponding to p-th target
The distance in the frequency domain is expressed as follows
Δδ ip ＝Δδ i -Δδ p
Step 2.4: after calculating and removing interference
Spectral peaks of spectral units
Directly making FFT on the intermediate frequency signal to obtain the spectrum peak S i,±0.5 ，S i,±0.5 The device consists of two parts: the ith target reflection signal is
The spectral peak obtained by FFT of the spectral unit and the p-th target reflection signal are FFT
Enhancement of spectral cells, from which
Expression (2)
Step 2.5: using spectral peaks
Calculating a deviation correction factor
Deviation from the spectral unit
Spectral peaks from step 2.4
Can be approximated as being present in the reflected signal only at the ith target
Spectral peaks of the spectral unit. Thus, a single target frequency estimation (Iterative frequency estimation by interpolation on Fourier coeffients [ J ] may be utilized]IEEE Transactions on Signal Processing,2005,53 (4): 1237-1242.) method to obtain the offset correction factor for the ith target
Deviation from the spectral unit
The expression is as follows
Step 2.5: q ← Q +1, judging whether Q is less than Q, if Q is less than or equal to Q, repeating steps 2.2-2.5
And updating Q after the current iteration frequency is Q each time of iteration is finished, and judging the relation between Q and the set iteration frequency Q.
And step 3: using spectral cells m i And spectral cell deviation
Calculating the frequency f of the ith target reflected signal i And the distance value R of the target i
When obtaining the deviation of the spectrum unit
Then, the real frequency spectrum unit corresponding to the ith target reflection signal
Thereby utilizing
Calculating a frequency value f corresponding to the ith target i From f i Obtaining the distance value R of the ith target by the relation of the distance i The expression is as follows
Wherein f is s Is the AD sampling frequency, c is the speed of light, B is the bandwidth of the modulation, T is the period of the modulated triangular wave.
To verify the performance of the algorithm, three typical methods were chosen For comparison, modified local-amplified Fourier algorithm (Zoom-FFT) (see the modified ZFT algorithm For improving LFMCW radar ranging accuracy, 2015, (2): 1-4.), non-iterative Frequency Estimation Method based on discrete Fourier transform (Candan algorithm) (see A Method For Fine Resolution Frequency Estimation From DFT Samples [ J ]. IEEE Signal Processing Letters,2011,18 (6): 351-354.), and spectral leakage suppression algorithm based on Hamming window (Hanming-window algorithm) (see the Fine Resolution Estimation From Samples: e of wired data [ J ]. Signal Processing, 114). The simulation effects of distance Root Mean Square Error (RMSE) under different Signal-to-Noise Ratio (SNR) conditions, RMSE of different algorithms under multiple targets and the like are mainly given.
The effect of the invention can be illustrated by the following simulations:
simulation conditions and contents:
simulation experiment using c 24.125GHz, bandwidth B250 MHz, signal period T8 us, sampling rate 95KHz, and FFT point number N512. Two targets in FIGS. 2 and 3Signal intensity of (b) satisfies A 2 ＝0.5A 1 . The signal strength of the two targets in FIG. 4 satisfies A 2 ＝0.5A 1 And SNR =15dB. Table 1 shows that the signal intensity varies in the range of A.epsilon. [0.3,1]Where a =1 indicates no signal loss on target reflection. The simulation experiments all adopt 5000 Monte-Carlo simulations.
As can be seen from fig. 2, the algorithm proposed herein has better estimation performance at the same SNR than the other three algorithms. Furthermore, the Zoom-FFT algorithm RMSE curve tends to be flat because the estimation accuracy of the algorithm is limited by the local refinement factor, resulting in the deviation value of the distance not varying with SNR.
As can be seen from fig. 3, the offset estimation of the target 2 by the proposed algorithm and the Candan algorithm has better performance, because both algorithms estimate the offset by the spectral peaks near the target, the peak values of the left and right adjacent spectral peaks caused by the interference of the target 1 are almost the same, and the offset correction factor is almost the same
Is less subject to peak variations, and therefore does not affect the bias
Therefore, the estimation performance of the two algorithms is better. In addition, the Hanming-window algorithm affects the estimation of the bias correction factor under the interference of the target 1, resulting in the failure of the distance estimation.
Fig. 4 is a graph of the effect of the number of signal sample points for target 2 on RMSE. From fig. 2, it can be seen that the Hanming-window algorithm and the Zoom-FFT algorithm have large error in the estimation of the bias under the condition, and therefore the Candan algorithm is selected for comparison. As can be seen from fig. 4, the simulation results of target 1 and target 2 are the same, so that only the simulation result of target 2 is given. As can be seen from the figure, the algorithm proposed herein can achieve smaller RMSE with a smaller number of iterations Q, and the RMSE is smaller with the same number of sample points.
Table 1 shows the simulation results for 10 targets whose signal strength ranges are a e [0.3,1], where a =1 indicates no signal loss when the target is reflected; it can be seen from the table that the signal amplitude has little influence on the estimation performance of the algorithm proposed herein, and in addition, it can be seen from the table that the distance difference between the targets 6, 7, 8 is less than 0.5m, and the estimation error of the algorithm for each target is small because the algorithm suppresses the influence of the spectral interference on the bias estimation in the iterative process.
TABLE 1 simulation parameters and results for multiple targets
FIG. 5 is an RMSE for various algorithms simulating the targets in Table 1. As can be seen from the figure, the estimation error of the algorithm proposed herein is stable below 10cm, the estimation error is more uniform and the RMSE is smaller compared with the other two algorithms, and it is found by combining the signal strength in table 1 that the estimation deviation of the algorithm proposed herein is less affected by the signal strength, because the spectrum change caused by the signal interference can be suppressed by the algorithm proposed herein, and the distance error estimated by the spectrum without interference is smaller. It can also be seen from the figure that the Zoom-FFT algorithm estimation error appears as a random error because the interpolation mode of Zoom-FFT is a uniform interpolation, the error is smaller as the spectral peak of the target is closer to the interpolation point, and the error is the largest when the spectral peak of the target just falls between the interpolation points.
Fig. 6 and 7 show the result of partially amplifying the targets 3 and 8, and the algorithms used are the algorithm proposed herein and the Zoom-FFT algorithm, and it can be seen from the figures that the accuracy can be improved by changing the number of iterations Q, and compared with the Zoom-FFT algorithm, the proposed algorithm is closer to the real frequency and has higher accuracy.
FIGS. 8 and 9 are actual environment test charts and FMCW radar hardware systems, respectively, with target 1 held stationary at 14.6m and target 2 at [8,10 ]]In the meantime, the test results of the FMCW radar for two target detection and the true values of the targets are shown in FIG. 10, where R is shifted every 0.2m real Representing the true distance value, R, of the object est Is the target distance estimated using the algorithm presented herein.
As can be seen from fig. 10, the FMCW radar output results are very close to the true range of the target and the error remains within 10 cm. In addition, FMCW radar data can be updated 20 times per second, proving that the algorithm fully meets the requirements of real-time.
The specific implementation mode of the invention also comprises the following steps:
step 1: performing Fast Fourier Transformation (FFT) on the intermediate frequency signal after AD sampling, and then performing Constant False Alarm Rate (CFAR) to estimate the number P of targets and the spectrum unit m where the targets are located i ,i＝1,2...P。
And 2, step: calculating the spectral unit m of the ith target i Corresponding deviation of spectral unit
i＝1,2…P。
Step 2.1: setting a target iteration number Q, and enabling the current iteration number Q =0; deviation estimate for ith target after single iteration
And amplitude estimator
Carry out initialization to
i＝1,2…P。
Step 2.2: estimating the reflected signal strength of the p-th target
p＝1,2...P；
Step 2.3: by using
Calculating the frequency spectrum interference generated by the p target to the i target
Step 2.4: after calculating and removing interference
Spectral peaks of spectral units
Step 2.5: using spectral peaks
Calculating a deviation correction factor
Deviation from the spectral unit
Step 2.6: q ← Q +1, judging whether Q is smaller than Q, if Q is not larger than Q, repeating steps 2.2-2.5.
And step 3: using spectral units m i And spectral cell deviation
Calculating the frequency f of the ith target reflected signal i And the distance value R of the target i
The expression of performing fourier transform on the intermediate frequency signal after AD sampling in the processing of step 1 is as follows
Wherein s is b (N) represents the intermediate frequency signal after AD sampling, N represents the number of points of FFT conversion, S b (k) Representing the amplitude of the k-th element in the frequency domain. The expression for estimating the target number P is as follows
P＝Count{S b (k)＞CFAR(S b (k))}
Wherein CFAR (S) b (k) Represents a pair S b (k) The result of the constant false alarm detection, count represents the S pair b (k) Greater than CFAR(S b (k) The number of the amplitude values is counted, and the counted result is the number of the targets.
The spectral unit m where the p-th target is located p The expression is as follows
In the step 2.1, the iteration number Q is set to be 2, experiments show that when Q is 2, high distance precision can be achieved, and the distance estimation precision is slightly influenced when multiple iterations are carried out.
Reflected signal strength of the p-th target in step 2.2
Is expressed as follows
Wherein
Is the estimated value of the intensity of the reflected signal of the first target, and the initial value is 0, Δ m pl Indicating the frequency spectrum unit m corresponding to the p-th target p Spectral unit m corresponding to the ith target l The distance in the frequency domain is expressed as follows
Δm pl ＝Δm p -Δm l
Δδ pl Indicating the deviation of the spectrum unit corresponding to the p-th target
Deviation of spectral unit corresponding to the l-th target
The distance in the frequency domain is expressed as follows
Δδ pl ＝Δδ p -Δδ l
Step 2.3 the p-th target generates spectrum interference to the i-th target
The expression is as follows
Wherein Δ m ip Indicating the frequency spectrum unit m corresponding to the ith target i M of the spectrum unit corresponding to the p-th target p The distance in the frequency domain is expressed as follows
Δm ip ＝Δm i -Δm p
Δδ ip Indicating the deviation of the spectrum unit corresponding to the ith target
Deviation of spectral unit corresponding to p-th target
The distance in the frequency domain is expressed as follows
Δδ ip ＝Δδ i -Δδ p
After the interference is removed in step 2.4
Spectral peaks of spectral units
The expression is as follows
Wherein S i,±0.5 Representation of direct FFT acquisition of an intermediate frequency signal
Spectral peak of spectral unit, S i,±0.5 The device consists of two parts: the ith target reflection signal is
The spectral peak obtained by FFT of the spectrum unit and the frequency spectrum interference of the p target to the i target are generated, and the expression is as follows
Deviation correction factor in step 2.5
Deviation from the spectral unit
The expression is as follows
Frequency f of ith target reflection signal in step 3 i And the distance value R of the target i The expression is as follows
Wherein f is s Is AD sampling frequency, c is light speed, B is frequency modulation bandwidth, T is frequency modulation triangle wave period
The invention provides an improved multi-target distance estimation algorithm based on iterative interpolation, aiming at the problem that the distance measurement precision of an on-vehicle anti-collision radar in a multi-target environment is easily interfered by frequency spectrums. The method comprises the steps of firstly estimating the number of targets and the rough position of the targets by utilizing fast Fourier transform and combining a constant false alarm detection algorithm, then carrying out non-uniform iterative interpolation on the frequency spectrum of the targets, and eliminating interference between the frequency spectrums in the process of each iteration, thereby obtaining more accurate distance. The algorithm provided by the simulation verification has higher estimation accuracy in a multi-target environment and under the condition of low signal to noise ratio. And finally, the algorithm is realized by DSP, and the algorithm is proved to meet the requirement of real-time property and is easy to realize in engineering.
Claims (8)
1. An FMCW radar distance estimation method based on fast iterative interpolation is characterized by comprising the following steps:
step 1: fourier transform is carried out on the intermediate frequency signals after AD sampling, and then constant false alarm detection is carried out to estimate the number P of targets and the frequency spectrum unit m where the targets are located i ，i＝1，2...P；
Step 2: calculating the spectral unit m of the ith target i Corresponding deviation of spectral unit
P, step 2 comprises the steps of:
step 2.1: setting a target iteration number Q, and enabling the current iteration number Q =0; deviation estimate for ith target after single iteration
And amplitude estimator
Proceed initialization to enable
Step 2.2: estimating the reflected signal strength of the p-th target
p＝1，2...P；
Step 2.3: by using
Calculating the frequency spectrum interference generated by the p target to the i target
Step 2.4: after calculating and removing interference
Spectral peaks of spectral units
Step 2.5: using spectral peaks
Calculating a deviation correction factor
Deviation from the spectral unit
Step 2.6: q ← Q +1, judging whether Q is smaller than Q, if Q is smaller than or equal to Q, repeating the steps 2.2-2.5; if Q is more than Q, executing step 3;
and 3, step 3: using spectral cells m i And spectral cell deviation
Calculating the frequency f of the ith target reflected signal i And the distance value R of the target i 。
2. The FMCW radar distance estimation method based on fast iterative interpolation of claim 1, wherein: in step 1, the intermediate frequency signal after AD sampling is subjected to fourier transform expression as follows:
wherein s is b (N) represents an intermediate frequency signal after AD sampling, N represents the number of points of Fourier transform, S b (k) And expressing the amplitude of the kth unit in the frequency domain, and estimating that the expression of the target number P satisfies the following conditions:
P＝Count{S b (k)＞CFAR(S b (k))}
wherein CFAR (S) b (k) Represents a pair S b (k) The result of the constant false alarm detection, count represents the S pair b (k) Greater than CFAR (S) b (k) The number of the amplitude values is counted, and the counted result is the number of the targets;
the spectral unit m where the p-th target is located p The expression is as follows:
3. an FMCW radar distance estimation method based on fast iterative interpolation according to claim 1, characterized in that: the number of iterations Q in step 2.1 is set to 2.
4. An FMCW radar distance estimation method based on fast iterative interpolation according to claim 1, characterized in that: reflected signal strength of the p-th target in step 2.2
The estimated value expression of (c) satisfies:
wherein
Is the intensity of the reflected signal of the ith targetDegree estimate, with an initial value of 0, N representing the number of points of the Fourier transform, Δ m pl Indicating the frequency spectrum unit m corresponding to the p-th target p Spectral unit m corresponding to the ith target l At a distance in the frequency domain, the expression satisfies:
Δm pl ＝Δm p -Δm l
Δδ pl indicating the deviation of the spectrum unit corresponding to the p-th target
Deviation of spectral unit corresponding to the l-th target
The distance in the frequency domain is expressed as follows:
Δδ pl ＝Δδ p -Δδ l 。
5. the FMCW radar distance estimation method based on fast iterative interpolation of claim 1, wherein: step 2.3, the p-th target generates spectrum interference to the i-th target
The expression satisfies:
where N represents the number of Fourier transform points, Δ m ip Indicating the frequency spectrum unit m corresponding to the ith target i M of the spectrum unit corresponding to the p-th target p At a distance in the frequency domain, the expression satisfies:
Δm ip ＝Δm i -Δm p
Δδ ip indicating the deviation of the spectrum unit corresponding to the ith target
Deviation of spectral unit corresponding to p-th target
At a distance in the frequency domain, the expression satisfies:
Δδ ip ＝Δδ i -Δδ p 。
6. the FMCW radar distance estimation method based on fast iterative interpolation of claim 1, wherein: after the interference is removed in step 2.4
Spectral peaks of spectral units
The expression satisfies:
wherein S i，±0.5 Representation obtained by direct Fourier transform of intermediate frequency signal
Spectral peak of spectral unit, S i，±0.5 The device consists of two parts: the ith target reflected signal is
The spectral peak and the p-th target obtained by Fourier transform of the frequency spectrum unit generate frequency spectrum interference on the i-th target, and the expression meets the following requirements:
wherein s is b (N) represents the intermediate frequency signal after AD sampling, and N represents the number of points of fourier transform.
7. The FMCW radar distance estimation method based on fast iterative interpolation of claim 1, wherein: deviation correction factor in step 2.5
Deviation from the spectral unit
The expressions respectively satisfy:
where N represents the number of points of the fourier transform.
8. An FMCW radar distance estimation method based on fast iterative interpolation according to claim 1, characterized in that: frequency f of ith target reflection signal in step 3 i And the distance value R of the target i The expressions respectively satisfy:
wherein f is s Is the AD sampling frequency, c is the speed of light, B is the bandwidth of modulation, T is the period of the triangular wave of modulation, N represents the number of points of Fourier transform.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811310159.8A CN109471095B (en) | 2018-11-06 | 2018-11-06 | FMCW radar distance estimation method based on fast iterative interpolation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811310159.8A CN109471095B (en) | 2018-11-06 | 2018-11-06 | FMCW radar distance estimation method based on fast iterative interpolation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109471095A CN109471095A (en) | 2019-03-15 |
| CN109471095B true CN109471095B (en) | 2023-02-14 |
Family
ID=65666960
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811310159.8A Active CN109471095B (en) | 2018-11-06 | 2018-11-06 | FMCW radar distance estimation method based on fast iterative interpolation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109471095B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109581052A (en) * | 2018-11-10 | 2019-04-05 | 中国人民解放军陆军勤务学院 | A kind of reality of iterated interpolation answers conversion frequency estimation method |
| CN111103588B (en) * | 2019-12-31 | 2022-12-13 | 哈尔滨工程大学 | Triangular wave multi-target identification method by utilizing signal energy |
| CN111337917B (en) * | 2020-03-17 | 2023-04-28 | 重庆邮电大学 | FMCW radar high-precision distance estimation method based on variable step interpolation iteration |
| CN111505618B (en) * | 2020-04-14 | 2022-06-24 | 东南大学 | Decoupling correction ranging method based on frequency estimation and suitable for field of vehicle-mounted millimeter wave radar |
| CN111521968B (en) * | 2020-05-22 | 2022-05-20 | 南京理工大学 | Underdetermined DOA estimation method based on target space diversity |
| CN112526474B (en) * | 2020-11-23 | 2022-05-27 | 哈尔滨工程大学 | FMCW radar range-velocity joint estimation method based on full-phase Fourier transform |
| CN112731307B (en) * | 2020-12-19 | 2023-01-03 | 湖北工业大学 | RATM-CFAR detector based on distance-angle joint estimation and detection method |
| CN113835077B (en) * | 2021-11-23 | 2022-02-11 | 中国空气动力研究与发展中心计算空气动力研究所 | Search radar target detection method and system based on variable pulse repetition frequency |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102608606A (en) * | 2012-03-22 | 2012-07-25 | 河海大学 | Effective method for accurately detecting multiple targets by automobile blind zone monitoring radar |
| CN103823216A (en) * | 2014-03-10 | 2014-05-28 | 中国科学院电子学研究所 | Distance measurement method for frequency modulation continuous wave radar system |
| CN103823215A (en) * | 2014-03-03 | 2014-05-28 | 中国科学院电子学研究所 | Linear frequency modulation continuous wave radar distance measuring method |
| KR101446439B1 (en) * | 2013-07-30 | 2014-10-06 | 재단법인대구경북과학기술원 | Method and device for estimating high-accuracy frequency |
| DE102013210256A1 (en) * | 2013-06-03 | 2014-12-04 | Robert Bosch Gmbh | INTERFERENCE SUPPRESSION ON AN FMCW RADAR |
| CN108415010A (en) * | 2018-04-28 | 2018-08-17 | 西安电子科技大学 | A kind of Radar Multi Target detection method based on trapezoidal LFMCW modulation |
| KR20180113238A (en) * | 2017-04-05 | 2018-10-16 | 주식회사화신 | Apparatus and method for assuming parameter of fmcw radar system |
-
2018
- 2018-11-06 CN CN201811310159.8A patent/CN109471095B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102608606A (en) * | 2012-03-22 | 2012-07-25 | 河海大学 | Effective method for accurately detecting multiple targets by automobile blind zone monitoring radar |
| DE102013210256A1 (en) * | 2013-06-03 | 2014-12-04 | Robert Bosch Gmbh | INTERFERENCE SUPPRESSION ON AN FMCW RADAR |
| KR101446439B1 (en) * | 2013-07-30 | 2014-10-06 | 재단법인대구경북과학기술원 | Method and device for estimating high-accuracy frequency |
| CN103823215A (en) * | 2014-03-03 | 2014-05-28 | 中国科学院电子学研究所 | Linear frequency modulation continuous wave radar distance measuring method |
| CN103823216A (en) * | 2014-03-10 | 2014-05-28 | 中国科学院电子学研究所 | Distance measurement method for frequency modulation continuous wave radar system |
| KR20180113238A (en) * | 2017-04-05 | 2018-10-16 | 주식회사화신 | Apparatus and method for assuming parameter of fmcw radar system |
| CN108415010A (en) * | 2018-04-28 | 2018-08-17 | 西安电子科技大学 | A kind of Radar Multi Target detection method based on trapezoidal LFMCW modulation |
Non-Patent Citations (1)
| Title |
|---|
| 基于快速迭代插值的FMCW雷达距离估计算法;段永昌等;《通信学报》;20190930;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109471095A (en) | 2019-03-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109471095B (en) | FMCW radar distance estimation method based on fast iterative interpolation | |
| CN104297740B (en) | Method for estimating Doppler spectrum of radar target on basis of phase analysis | |
| KR101294681B1 (en) | Apparatus and method for processing weather signal | |
| CN108169739B (en) | Linear frequency modulation continuous wave time-width ratio estimation method based on fractional Fourier transform and minimum pulse width detection | |
| CN103616687A (en) | Polynomial fitting ISAR envelope alignment method based on piecewise linear estimation | |
| Vizitiu et al. | The synthesis of some NLFM laws using the stationary phase principle | |
| CN103760534A (en) | Method for correcting ionized layer chromatic dispersion effect of satellite borne SAR data | |
| CN104155653B (en) | SAR back projection imaging method based on feature distance subspace | |
| CN109061599B (en) | STAP method based on cyclostationarity and symmetric prior knowledge | |
| CN114545342A (en) | Radar pulse signal parameter measuring method using multi-channel reconnaissance receiver | |
| CN107861115B (en) | OTHR maneuvering target parameter estimation method based on instantaneous autocorrelation matrix sparse decomposition | |
| CN112130138B (en) | Ionized layer phase pollution correction method for over-the-horizon radar | |
| CN110658517B (en) | Dimensionality reduction sparse STAP method and device based on uncertain priori knowledge | |
| CN115826004B (en) | Three-star cooperative direct positioning method based on two-dimensional angle and time difference combination | |
| CN107315169B (en) | Clutter covariance matrix estimation method based on second-order statistic similarity | |
| Torres et al. | The impact of signal processing on the range-weighting function for weather radars | |
| CN111551931B (en) | Target length feature extraction method based on polarization high-resolution range profile | |
| CN111007473B (en) | High-speed weak target detection method based on distance frequency domain autocorrelation function | |
| CN116996137B (en) | Low signal-to-noise ratio broadband linear frequency modulation signal detection method based on weighted superposition | |
| Deng et al. | Doppler rate estimation on coherent sinusoidal pulse train and its Cramer–Rao lower bound | |
| CN111308436A (en) | Radar space-time adaptive processing method and device based on volume correlation function | |
| CN116449304B (en) | SAR emission pulse arrival time measurement method based on frequency measurement | |
| CN113702969B (en) | Micro Doppler signal parameter estimation method based on self-adaptive STFT method | |
| CN111948451B (en) | Frequency modulation continuous wave signal direction finding method based on cross-spectrum phase weighted average method | |
| CN114609606B (en) | Parameter estimation method and device of target object and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |