High-speed multi-target distance and speed joint estimation method, device and program product

Abstract

The invention discloses a high-speed multi-target distance and speed joint estimation method, device and program product, relates to the field of radar signal processing, and is used for obtaining a stable and reliable multi-target distance and speed joint estimation result. FRFT processing is carried out on echo signals, peak points are extracted from an amplitude matrix of a processing result, effective peak points are screened out, effective target energy rays are extracted according to the effective peak points of each row, smooth filtering processing is carried out on the effective target energy rays, and the effective target energy rays are obtained. And finally, estimating the distance and the speed of the corresponding target according to the angle sampling value and the u-axis projection value of the effective peak point with the maximum amplitude on each effective target energy line. According to the method, the influence of the target speed on the radar echo frequency modulation slope is considered, the estimation result is more accurate, the joint estimation of the distance and the speed is realized on the single-cycle echo, and the average effect of multiple cycles is avoided.

Description

Technical Field

[0001] The invention relates to the field of radar signal processing, and in particular to a high-speed multi-target distance and speed joint estimation method based on FRFT. Background Art

[0002] Linear Frequency Modulation (LFM) pulse radar is a major means of space target detection, parameter estimation, and tracking, but there are many challenges in parameter estimation of high-speed multi-targets in space. For high-speed targets in space, there will be a significant range-velocity coupling effect in the LFM pulse radar echo, and it is difficult to achieve range-velocity joint estimation from a single-cycle echo using conventional means.

[0003] The traditional method is to use multi-period radar echoes, extract the target peak position after pulse compression, and estimate the target distance and speed by establishing a target motion model. However, the positions and speeds of multiple targets are independent of each other, and there may be complex features such as overlapping target trajectories and changes in motion and maneuvering, which will increase the error of target distance and speed parameter estimation, or even make it invalid.

[0004] Some scholars have also proposed other methods to solve the problem of joint estimation of target distance and speed. The Chinese patent document with publication number CN112526474A proposed a "FMCW radar distance and speed joint estimation method based on full-phase Fourier transform", which uses full-phase Fourier transform to achieve joint estimation of target distance and speed. The Chinese patent document with publication number CN118011381A proposed a "distance and speed joint estimation method, system and medium based on FMCW radar", which realizes the joint estimation of target distance and speed through 2D-FFT transformation and CZT algorithm. However, both methods do not take into account the change of radar echo frequency modulation slope caused by the different speeds of multiple targets, which leads to theoretically inevitable errors in the distance and speed estimation of high-speed or ultra-high-speed targets, and loses the accuracy of estimation. In addition, both methods need to use multiple cycles of radar echoes for processing. Due to the existence of averaging effect, they cannot adapt well to complex features such as target maneuvers. Summary of the invention

[0005] The object of the present invention is to provide a high-speed multi-target distance and speed joint estimation method, device and program product to obtain robust and reliable multi-target distance and speed joint estimation results in order to address all or part of the above-mentioned problems.

[0006] The technical solution adopted by the present invention is as follows:

[0007] A high-speed multi-target distance and speed joint estimation method based on FRFT (fractional Fourier transform), comprising:

[0008] Down-convert and sample the radar echo signal to obtain a digital echo signal;

[0009] Performing FRFT processing on the digital echo signal to obtain a processing result;

[0010] Obtaining an amplitude matrix of the processing result, filtering the amplitude matrix according to a set threshold, and extracting peak points for each row;

[0011] According to the set target minimum spacing, the peak points of each row are screened respectively to obtain the retained valid peak points;

[0012] Extracting effective target energy lines according to the effective peak points of each row, and performing smoothing filtering on the effective target energy lines;

[0013] According to the angle sampling value and the u-axis projection value of the effective peak point with the largest amplitude on each of the effective target energy lines, the distance and speed of the corresponding target are estimated.

[0014] The present application also provides a device, which includes a processor and a storage medium, wherein the storage medium stores a computer program, and when the processor runs the computer program in the storage medium, it executes the above-mentioned FRFT-based high-speed multi-target distance and speed joint estimation method.

[0015] The present application also provides a computer program product, including a computer program, which, when executed by a processor, executes the above-mentioned FRFT-based high-speed multi-target distance and speed joint estimation method.

[0016] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0017] The high-speed multi-target distance and speed joint estimation method based on FRFT proposed in this application takes into account the influence of target speed on the frequency modulation slope of radar echo, making the echo model more accurate. In addition, this application realizes the joint estimation of distance and speed on a single-cycle echo, avoiding the averaging effect of multiple cycles, and can adapt well to complex characteristics such as target maneuverability, thereby obtaining a robust and reliable multi-target distance-speed joint estimation result. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The present invention will now be described by way of example with reference to the accompanying drawings, in which:

[0019] Figure 1 It is a flow chart of a high-speed multi-target distance and speed joint estimation method based on FRFT provided in an embodiment of the present application.

[0020] Figure 2 This is a simulation diagram of the FRFT processing amplitude when searching for the first angle in an embodiment of the present application.

[0021] Figure 3 This is a simulation diagram of the FRFT processing amplitude for all search angles in the embodiment of the present application.

[0022] Figure 4 It is a simulation diagram of the relationship curve between the amplitude and speed of all target energy lines in the embodiment of the present application. DETAILED DESCRIPTION

[0023] All features disclosed in this specification, or steps in all methods or processes disclosed, except mutually exclusive features and / or steps, can be combined in any manner.

[0024] Any feature disclosed in this specification (including any additional claims and abstract), unless otherwise stated, may be replaced by other equivalent or alternative features having similar purposes. That is, unless otherwise stated, each feature is only an example of a series of equivalent or similar features.

[0025] The conventional method of using multi-period radar echoes for target distance-velocity joint estimation has large estimation errors, and the prior art does not take into account the inevitable error caused by the change in the frequency modulation slope of the radar echo caused by the different speeds of multiple targets, and the prior art needs to use radar echoes of multiple periods to introduce the averaging effect. The embodiments of the present application provide a high-speed multi-target distance and velocity joint estimation method, device and program product, aiming to obtain a robust and reliable multi-target distance and velocity joint estimation result, avoiding the multi-period averaging effect.

[0026] Description of common knowledge that may be involved in this application:

[0027] 1) Linear frequency modulation signal.

[0028] The expression of linear frequency modulation signal s(t) is:

[0029] s(t)=G(t)exp(j(2πf b t+πkt 2 )), (1-1)

[0030] Where t is the time variable, exp(·) is the exponential function, j is the imaginary unit, and f b is the starting frequency of the linear frequency modulation signal, is the frequency modulation slope of the linear frequency modulation signal, B is the transmission signal bandwidth, P is the transmission signal pulse width, G(t) is the amplitude envelope window function, G(t) can be taken as a rectangular window function, and the expression is:

[0031]

[0032] 2)FRFT.

[0033] The definition of FRFT is:

[0034]

[0035] Among them, x(t) is the signal to be analyzed, α is the angle, u is the displacement factor, α and u together form a two-dimensional transformation domain, K α (t,u) is the transformation kernel of FRFT, and its expression is:

[0036]

[0037] Where δ(t) represents a unit impulse signal, and q is an arbitrary integer. It should be noted that in equation (1-4), the cases of α=2qπ ​​and α=(2q+1)π are only to ensure the rigor of the fractional Fourier transform expression. In practical applications, the value range of α is usually set to: That is, α is in the first or fourth quadrant of the two-dimensional time-frequency domain. For the case where α is in the second or third quadrant of the time-frequency domain, it can be easily equivalent to the fourth or first quadrant, which will not be discussed in detail here.

[0038] In some embodiments of the present application, a high-speed multi-target distance and speed joint estimation method based on FRFT includes the following process:

[0039] S1. Down-convert and sample the radar echo signal to obtain a digital echo signal.

[0040] At the radar receiving end, for each pulse period (PRI), the radar echo signal r(t) is received. As a feasible implementation method, the radar echo signal is down-converted and digitally sampled to obtain a digital echo signal r d (n), which is expressed as:

[0041] r d (n) = r(nT s ), (1-5)

[0042] Among them, T s is the sampling interval, and its reciprocal is the sampling rate f s =1 / T s , n is the index value of the discrete sequence, and its value is: 0,1,2,...,N r -1, where N rFor r d (n) The total number of sampling points.

[0043] S2, digital echo signal r d (n) Perform FRFT processing to obtain a processing result.

[0044] As a feasible implementation, the above-mentioned digital echo signal r d (n) performing FRFT processing, including:

[0045] S2.1, according to the set target speed change interval [v d ,v u ] and the speed search step Δv, calculate the angle sampling value α required for FRFT processing i The calculation process includes:

[0046] According to the set target speed change interval [v d ,v u ] and the speed search step Δv to obtain the speed sampling value v i :

[0047] v i =v d +iΔv, (1-6)

[0048] Where i = 0, 1, 2, ..., M-1, where M is the total number of target speeds to be searched, and its value is:

[0049]

[0050] Here, round(·) means rounding to the nearest integer.

[0051] According to the speed sampling value v i , calculate the corresponding speed proportional coefficient η i :

[0052]

[0053] Where c represents the speed of light, and its value can be taken as c = 3 × 10 8 m / s.

[0054] Then the angle sampling value α to be searched in the FRFT transform i Calculate as follows:

[0055]

[0056] Wherein, arctan(·) is the inverse tangent function, k is the frequency modulation slope of the linear frequency modulation signal s(t) as mentioned above, and i=0, 1, 2, ..., M-1.

[0057] S2.2. Calculate the u-axis sampling interval u required for FRFT processing according to the set radial distance resolution interval Δl s .

[0058]

[0059] Among them, cos(·) is the cosine function, It means traversing i∈{0,1,2,...,M-1} and taking the minimum value among them.

[0060] S2.3, according to the above angle sampling value α i and u-axis sampling interval u s , for the digital echo signal r d (n) Perform FRFT processing.

[0061] Step S2.1 obtains M angle sampling values ​​α i , plus the u-axis sampling interval u obtained in step S2.2 s , for the digital echo signal r d (n) Perform FRFT processing to obtain the processing results The processing results are arranged in discrete matrix form.

[0062]

[0063] in, is the FRFT transformation kernel function, and its value is calculated according to formula (1-4). p is the discretization index of the u-axis, and its value is: p=0,1,2,...,N-1, where N represents the total number of sampling points of the u-axis.

[0064] There are M rows and N columns in total, where the vertical row direction corresponds to the sampling value α of each angle i , i=0,1,2,...,M-1, the horizontal column direction corresponds to the u-axis projection value pu s , p=0,1,2,...,N-1.

[0065] S3. Get processing results The magnitude matrix According to the set threshold, the amplitude matrix Filter and extract the peak points for each row.

[0066] As a feasible implementation method, in step S3, the amplitude matrix is Filtering includes:

[0067] S3.1. Based on processing results The threshold is set by the average power of

[0068] Assume P noise express The average power of ζ is the amplitude multiple coefficient, satisfying ζ≥1. In some optional implementations, the threshold Thi is selected as follows:

[0069]

[0070] S3.2, the amplitude matrix The amplitudes smaller than the threshold value Thi are set to zero, and those larger than the threshold value Thi are retained.

[0071] In addition, step S3 also includes:

[0072] S3.3, the amplitude matrix Each row of data in the is processed by peak point extraction, and the peak point extraction method can be based on the existing technology. Record the extracted peak point corresponding to the amplitude matrix The angle sampling value α in i , u-axis projection value pu s and amplitude information.

[0073] S4, according to the set target minimum distance d min , filter the peak points of each row respectively to obtain the retained valid peak points.

[0074] As a feasible implementation, in step S4, the peak points of each row are screened according to the set target minimum spacing, including:

[0075] S4.1. According to the set target minimum distance d min , calculate the minimum projection spacing u projected onto the u axis min .

[0076] By the target minimum distance d min Convert the minimum projection distance u min The method is:

[0077]

[0078] S4.2. Minimum projection spacing u in each row min Only one peak point is retained within the range as a valid peak point.

[0079] In some embodiments, step S4.2 includes:

[0080] Based on the minimum projection distance u min Group the peak points of each row.

[0081] The peak point with the largest amplitude in each group is retained as the valid peak point.

[0082] The above grouping process can be as follows: judge each peak point of each row in turn, if the distance between the current peak point and the previous peak point on the u-axis is less than or equal to the minimum projection spacing u min , the current peak point is classified as the same group as the previous peak point, otherwise the current peak point is classified as the starting point of the next group. For the peak points in the same group, take the corresponding amplitude matrix |X αi (pu s )| is retained as a valid peak point, while the other peak points in the same group are discarded. According to this method, the amplitude matrix |X αi (pu s )|The peak points of each row are screened to obtain the amplitude matrix |X αi (pu s )|. These effective peak points record the corresponding angle sampling value α i , u-axis projection value pu s and amplitude information.

[0083] S4. Extract the effective target energy line according to the effective peak point of each row, and perform smoothing filtering on the effective target energy line.

[0084] In a feasible implementation manner, extracting effective target energy lines according to the effective peak points of each row in step S4 includes:

[0085] S4.1. Extract all target energy lines according to the effective peak points of each row.

[0086] In some specific implementations, the target energy line extraction process includes:

[0087] According to the effective peak points of each row, the closest effective peak points in different rows are sequentially associated with the same target energy line. Specifically, there are:

[0088] Record the index position information of the valid peak point in the first row;

[0089] For other rows, the valid peak point in the row that is closest to the valid peak point recorded in the previous row is sequentially associated with the valid peak point recorded in the previous row on the same target energy line.

[0090] According to this method, multiple target energy lines can be obtained, and each target energy line contains effective peak points in different rows.

[0091] S4.2. Filter out valid target energy lines from all target energy lines according to the set valid peak point number threshold.

[0092] For the target energy line obtained, it is necessary to judge its validity. As a feasible implementation method, the minimum number N of valid peak points on the target energy line is set. p , N p Satisfy 1≤N p ≤M. Determine in turn whether the number of effective peak points on each target energy line is greater than or equal to N p If the condition is met, it is a valid target energy line, otherwise it is an invalid target energy line. The number of valid target energy lines is the number of targets actually detected.

[0093] Step S4 also includes:

[0094] S4.3, smoothing and filtering the effective target energy line amplitude to obtain a smooth effective target energy line. The filter may be a digital window function filter.

[0095] S6. Estimate the distance s of the corresponding target based on the angle sampling value and u-axis projection value of the effective peak point with the largest amplitude on each effective target energy line. obj and speed v obj .

[0096] For each effective peak point, its corresponding angle sampling value α is recorded i , u-axis projection value pu s In step S6, for each valid target energy value, obtain the angle sampling value α of the valid peak point with the largest amplitude. obj and the u-axis projection value u obj , calculate the distance estimate s of the corresponding target obj and the velocity estimate v obj The calculation method is:

[0097]

[0098] Where cot(·) represents the cotangent function.

[0099] The embodiment of the present application performs a refined modeling of the LFM radar echo signal r(t), which is expressed as follows:

[0100]

[0101] Among them, L is the number of targets in the scene to be detected, A i ,θ i , τ i With ξ i are the amplitude, phase, delay, and scale change factor of the echo component corresponding to the i-th target, ξ i Satisfaction relationship i =1-η i ,in, v i is the target radial velocity.

[0102] It can be analyzed that according to the above refined echo model, the frequency modulation slopes of the echo components corresponding to different target speeds are also different, and the frequency modulation slope value is By using FRFT processing, the echo signal is projected into the (α,u) domain, where α is directly linked to the frequency modulation slope of the echo signal, and the u axis is directly related to the delay and frequency. By using the FRFT processing of the echo, the distance and speed information of multiple targets can be solved in a single cycle echo according to the method of this application. This reflects the timeliness of this application in the joint estimation of multi-target distance and speed, as well as the characteristics of adaptability to mobility.

[0103] The existing technology assumes that the echoes of targets with different speeds have the same frequency modulation slope, which makes it difficult to simultaneously calculate the distance and speed information of multiple targets from a single cycle. In order to obtain the distance and speed estimation results of multiple targets, it is necessary to combine the multi-cycle echoes for unified processing. Due to the averaging effect and inertia effect, the maneuverability of the target will cause model mismatch and increase in estimation error. In addition, the model error of the non-refined echo modeling will be significantly reflected when solving the motion parameters of high-speed or ultra-high-speed targets, resulting in increased parameter estimation deviation and echo distance image mismatch.

[0104] The device provided in the present application, in some embodiments, includes a processor and a storage medium, in which a computer program is stored. When the processor runs the computer program in the storage medium, the above-mentioned FRFT-based high-speed multi-target distance and speed joint estimation method is executed.

[0105] The computer program product provided by the present application, in some embodiments, includes a computer program, which, when executed by a processor, executes the above-mentioned high-speed multi-target distance and speed joint estimation method based on FRFT. As a feasible implementation method, the computer program product can be recorded and stored in a computer-readable storage medium, such as a CD, a USB flash drive, a disk, a hard disk, etc.

[0106] The embodiments of the present application also verify the performance of the proposed method.

[0107] 1) Simulation conditions

[0108] Assume that the scene to be detected contains three hypersonic targets, whose true radial distances are 96.7km, 97.0km, and 98.4km, respectively, and whose true radial speeds are 4.00 Mach, 9.25 Mach, and 7.5 Mach, respectively. The amplitude gains of the echo components corresponding to them are all set to 1. Assume that the radar transmitting signal is a broadband linear frequency modulation pulse signal with a signal bandwidth of 500MHz and a pulse width of 0.5ms. The modulation method in which the frequency increases linearly with time is adopted. The input signal-to-noise ratio of the echo is set to -25dB, the intermediate frequency sampling rate is set to 1.2GHz, and the total number of sampling points of the echo signal to be analyzed is 640000. The distance step of FRFT processing is set to 0.14m, and the speed step is set to 0.25 Mach, and a total of 60 angles α to be searched can be obtained. Simulation computer configuration: CPU 3.8GHz, memory 16GB. The other parameters that have no substantial impact on the simulation results are not listed here.

[0109] 2.) Simulation content

[0110] Under the above simulation conditions, a radar echo signal is constructed, and the echo signal is subjected to FRFT processing using the method provided in this application. The FRFT processing results are recorded as follows: Figure 2-Figure 4 shown.

[0111] Figure 2 The simulation results of the FRFT transform amplitude at the first search angle are given. It can be seen from the results that three target peaks can be clearly seen in the FRFT transform results, and the highest peak value of each target is accurately found according to the method of the present application. Figure 3 The simulation results of the two-dimensional FRFT transform amplitude in the velocity-distance dimension at all search angles are given, and the three target energy lines can be clearly distinguished. Figure 4 The curves of the amplitudes of the three target energy lines relative to the speed are given, and the speed value corresponding to the highest peak of each target energy line is accurately found according to the method of this application. Table 1 shows the target distance-speed joint estimation value of the method of this application and the comparison results with the true value. It can be seen from the results that the method of this application accurately realizes the distance-speed joint estimation of multiple targets, and the estimated distance and speed errors do not exceed the set distance and speed step values.

[0112] Table 1 Target distance-speed joint estimation value and comparison with the true value

[0113]

[0114] In addition, it takes 9.57 seconds for the computer to complete the entire simulation calculation. It should be noted that the FRFT calculations for each search angle are performed serially during the simulation. In actual embedded hardware applications, each search angle can be calculated in parallel, so the time consumption of the method of the present application in actual applications is sufficient to meet real-time requirements.

[0115] The present invention is not limited to the above-mentioned specific embodiments, but extends to any new features or any new combination disclosed in this specification, as well as any new method or process steps or any new combination disclosed.

Claims

1. A high-speed multi-target distance and velocity joint estimation method based on FRFT, characterized in that: include: Down-convert and sample the radar echo signal to obtain a digital echo signal; Performing FRFT processing on the digital echo signal to obtain a processing result; Obtaining an amplitude matrix of the processing result, filtering the amplitude matrix according to a set threshold, and extracting peak points for each row; According to the set target minimum spacing, the peak points of each row are screened respectively to obtain the retained valid peak points; Extracting effective target energy lines according to the effective peak points of each row, and performing smoothing filtering on the effective target energy lines; According to the angle sampling value and the u-axis projection value of the effective peak point with the largest amplitude on each of the effective target energy lines, the distance and speed of the corresponding target are estimated.

2. The high-speed multi-target distance and speed joint estimation method based on FRFT as claimed in claim 1, characterized in that: Performing FRFT processing on the digital echo signal includes: According to the set target speed change range and speed search step, calculate the angle sampling value required for FRFT processing; According to the set radial distance resolution interval, calculate the u-axis sampling interval required for FRFT processing; The digital echo signal is subjected to FRFT processing according to the angle sampling value and the u-axis sampling interval.

3. The high-speed multi-target distance and speed joint estimation method based on FRFT as claimed in claim 1, characterized in that: The filtering the amplitude matrix according to a predetermined threshold comprises: Setting the threshold based on the average power of the processing result; The amplitudes in the amplitude matrix that are smaller than the threshold are set to zero.

4. The high-speed multi-target distance and speed joint estimation method based on FRFT as claimed in claim 1, characterized in that: The step of screening the peak points of each row according to the set target minimum spacing comprises: According to the set target minimum spacing, calculate the minimum projection spacing projected to the u-axis; Only one peak point is retained as a valid peak point within the minimum projection spacing range in each row.

5. The high-speed multi-target distance and speed joint estimation method based on FRFT as claimed in claim 4, characterized in that: The retaining only one peak point as a valid peak point within the minimum projection spacing range in each row includes: Grouping the peak points of each row based on the minimum projection distance; The peak point with the largest amplitude in each group is retained as the valid peak point.

6. The high-speed multi-target distance and speed joint estimation method based on FRFT as claimed in claim 1, characterized in that: The step of extracting effective target energy lines according to the effective peak points of each row includes: Extract all target energy lines according to the effective peak point of each row; According to the set effective peak point number threshold, the effective target energy lines are screened out from all target energy lines.

7. The high-speed multi-target distance and speed joint estimation method based on FRFT as claimed in claim 6, characterized in that: The step of extracting all target energy lines according to the effective peak point of each row includes: According to the effective peak points of each row, the effective peak points closest to each other in different rows are sequentially associated with the same target energy line.

8. The high-speed multi-target distance and speed joint estimation method based on FRFT as claimed in claim 7, characterized in that: The step of sequentially associating the closest valid peak points in different rows to the same target energy line includes: Record the index position information of the valid peak point in the first row; For other rows, the valid peak point in the row that is closest to the valid peak point recorded in the previous row is sequentially associated with the valid peak point recorded in the previous row on the same target energy line.

9. A device comprising a processor and a storage medium, wherein the storage medium stores a computer program, characterized in that: When the processor runs the computer program in the storage medium, it executes the high-speed multi-target distance and speed joint estimation method based on FRFT as described in any one of claims 1-8.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the high-speed multi-target distance and speed joint estimation method based on FRFT as described in any one of claims 1-8 is executed.

Images