A high-speed target detection method and device for pulse Doppler radar

Abstract

The application discloses a high-speed target detection method for pulse Doppler radar, and is characterized in that the method comprises the following steps: step 1, obtaining echo signals to obtain initial data; step 2, resampling the initial data in a slow time dimension to obtain resampled data; step 3, compensating the resampled data through a velocity ambiguity number matching filter to obtain corresponding corrected data; step 4, performing linear transformation processing on the corrected data to obtain target data; step 5, detecting the target data to construct an effective target parameter set; and step 6, repeating the above steps to output a final effective target parameter set. The application further provides a high-speed target detection device. The method provided by the application can solve the problems of cross-range cells and velocity ambiguity of super-high-speed targets in multi-target detection in a complex scene, thereby realizing high-precision estimation and detection of the distance and velocity of super-high-speed multi-targets.

Description

Technical Field

[0001] This invention belongs to the field of radar signal processing technology, and particularly relates to a high-speed target detection method and device for pulse Doppler radar. Background Technology

[0002] Due to the rapid development of aviation and aerospace technologies, aircraft, missiles, artificial satellites, and spacecraft all employ radar as a means of detection and control. Especially after the successful development of anti-intercontinental ballistic missile systems, requirements for radar have been placed on long-range, high-precision, high-resolution, and multi-target detection. Simultaneously, high-speed multi-target radar detection is of great significance in target imaging and space exploration. Generally, for targets with aligned range images, moving target detection (MTD) algorithms can be used. However, in the field of ultra-high-speed target detection, low radar cross-section (RCS) is a major problem faced by modern radar systems. To address this challenge, many solutions propose extending the signal coherence integration time to improve the target signal-to-noise ratio. However, the high-speed movement of the target causes range drift and Doppler broadening, severely affecting the effective integration of echo energy and reducing detection performance.

[0003] Patent document CN115308738A discloses a swarm density estimation algorithm based on centimeter-wave scanning radar. By dividing the three-dimensional space into independent resolution units and then statistically analyzing the detection units of the detected target, the algorithm can better obtain the spatial distribution and three-dimensional shape of the swarm. By using the relationship between echo intensity and effective cross-sectional area of ​​target reflection in the radar power formula, the number of individuals in the detection unit is derived. Combined with the distance and velocity distribution map generated by the Doppler effect, the number of individuals in different detection units is verified, making the statistical analysis of the overall number of swarms in the airspace more reasonable and ensuring the accuracy of the final density value estimation.

[0004] Patent document CN114063057A discloses a multi-functional radar signal processing method for maritime surveillance. It adopts a DBF (Digital Doppler) system, which is mainly for searching and forms a three-dimensional coverage of the monitored area, and performs preliminary classification of surface and air targets. For stable sea surface targets, it distinguishes them according to the amplitude of the echo. Large targets are processed by step-frequency imaging, and the size of ships is distinguished according to the size of the target image. For air targets, it distinguishes low-altitude aircraft, birds, etc. according to the stability of the flight path and micro-Doppler characteristics. Summary of the Invention

[0005] The purpose of this invention is to provide a high-speed target detection method and device for pulse Doppler radar. This method can solve the problem of ambiguity in the range and velocity of ultra-high-speed targets when detecting multiple targets in complex scenarios, thereby enabling high-precision estimation and detection of the range and velocity of ultra-high-speed multiple targets.

[0006] To achieve the first objective of this invention, the following technical solution is provided: a high-speed target detection method for pulse Doppler radar, specifically comprising:

[0007] Step 1: Acquire the echo signal emitted by the pulse Doppler radar and perform linear processing on the echo signal to obtain the initial data of velocity and range coupling for each target.

[0008] Step 2: Resample the initial data in the slow time dimension to obtain resampled data after decoupling speed and distance;

[0009] Step 3: Compensate the resampled data using a pre-built velocity ambiguity number matched filter to obtain the corresponding corrected data;

[0010] Step 4: Perform a linear transformation on the corrected data to obtain the target data;

[0011] Step 5: Extract the first data in the distance dimension and the second data in the velocity dimension from the target data, and calculate the square of the amplitude modulus of the first and second data. Use the highest value of the square of the amplitude modulus as the input of the pre-constructed two-dimensional constant false alarm detector for detection.

[0012] If the target is detected as a valid target, the target's distance, velocity, and amplitude in the input target data are added to the valid target parameter set.

[0013] Conversely, the corresponding target data is removed.

[0014] Step 6: Repeat steps 1 to 5 until the termination condition of the two-dimensional constant false alarm detector is met, so as to output the final set of effective target parameters.

[0015] This invention corrects the distance migration problem of high-speed targets by resampling in the slow time dimension, thereby improving the accuracy of high-speed target distance estimation. It also solves the problem of low detection accuracy caused by the mutual influence of multiple strong and weak targets by eliminating strong targets. Finally, it improves the accuracy of high-speed target velocity estimation by constructing a matched filter to search and compensate for the velocity ambiguity number corresponding to the target.

[0016] Specifically, the pulse Doppler radar is a shore-based radar, an airborne radar, or a shipborne radar.

[0017] Specifically, in step 2, in the slow time dimension, satisfying t m =[f c / (f c +f)]τ m Resampling is performed under the condition of sinc interpolation, where τ m For virtual time, t m This is the original time.

[0018] Specifically, the expression for the resampling is as follows:

[0019]

[0020] Where m is the original sampling point, m' is the sampling point after sinc interpolation, f is the fast time frequency, and f c Where M is the carrier frequency and M is the number of pulses.

[0021] Specifically, the expression for the corrected data is as follows:

[0022]

[0023] in, τ represents the velocity ambiguity number used to compensate for resampled data. m For virtual time, This represents a velocity ambiguity number matched filter.

[0024] Specifically, in step 4, the linear transformation process is as follows:

[0025] The corrected data is subjected to inverse fast Fourier transform in the fast time dimension and fast Fourier transform in the slow time dimension to obtain the target data in the distance-Doppler dimension.

[0026] Specifically, in step 5, the detection process of the two-dimensional constant false alarm rate detector is as follows:

[0027] Remove guard cells from all training units that are near the maximum squared magnitude of the input.

[0028] Calculate the average of all data except for the maximum squared value of the amplitude magnitude to obtain the noise floor value;

[0029] A threshold for an effective target is determined based on the noise floor value, and the ratio of the maximum squared magnitude of the amplitude to the noise floor value is compared with the threshold.

[0030] If the ratio is greater than the threshold, then there is a valid target; otherwise, there is an invalid target.

[0031] Specifically, the termination conditions include:

[0032] a. The number of repetitions has reached the maximum number of iterations preset by the constant false alarm rate detector;

[0033] b. The constant false alarm rate detector determines that there is no target, that is, the detection statistic corresponding to the maximum value point does not exceed the corresponding threshold.

[0034] To achieve the second objective of this invention, the following technical solution is provided: a high-speed target detection device for performing the steps of the above-described high-speed target detection method for pulse Doppler radar.

[0035] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0036] Compared with traditional pulse compression methods based on matched filtering, this invention overcomes the distance migration problem and improves the distance estimation accuracy by correcting through resampling in the slow time dimension;

[0037] By constructing a matched filter for the velocity ambiguity number to search for and compensate for the influence of velocity ambiguity, the accuracy of velocity estimation for high-speed multi-target targets is further improved.

[0038] By removing the signal data of the estimated strong targets before each detection, the probability of successfully detecting weak targets is greatly increased. Attached Figure Description

[0039] Figure 1 This is a flowchart of a high-speed target detection method for pulse Doppler radar provided in this embodiment;

[0040] Figure 2 This is a schematic diagram of the structure of a high-speed target detection device provided in this embodiment;

[0041] Figure 3 The detection result of target one using the high-speed target detection method provided in this embodiment;

[0042] Figure 4 The detection result of target two using the high-speed target detection method provided in this embodiment;

[0043] Figure 5 The detection result of target three using the high-speed target detection method provided in this embodiment;

[0044] Figure 6 The velocity ambiguity number detection results for target one, target two, and target three provided in this embodiment;

[0045] Figure 7 The detection result of target one using the conventional pulse compression method provided in this embodiment;

[0046] Figure 8 The detection result of target two using the conventional pulse compression method provided in this embodiment;

[0047] Figure 9 The detection results of target three using the traditional pulse compression method provided in this embodiment. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0049] like Figure 1 As shown, this embodiment provides a high-speed target detection method for pulse Doppler radar, which includes:

[0050] Step 1: Receive the echo signal transmitted by the pulse Doppler radar and obtain the velocity and range coupled data through linear processing:

[0051] First, initialize the two-dimensional constant false alarm rate (CFAR) detector by inputting the false alarm probability, maximum number of iterations, oversampling factor, correction count, number of guard units, and number of training units. Then, obtain K target echo signals from the pulse Doppler radar via down-conversion demodulation. Expressed as a formula:

[0052]

[0053] Where N = F s T,r m,k =r 0,k +mTv k ,m=0,1,…,M-1, where m is the radial distance of the k-th target from the radar when the m-th pulse is transmitted; F s T is the sampling frequency. s =1 / F s f is the sampling time. c Where c is the carrier frequency, c is the speed of light, T is the pulse repetition period, and v is the carrier frequency. k Let be the velocity of the k-th target.

[0054] The speed of high-speed targets often exceeds the maximum unambiguous speed of radar systems, thus causing speed ambiguity. That is, the target speed can be expressed as an equation concerning the speed ambiguity number, as follows:

[0055] v k =γ k v amb +v k,dect

[0056] Among them, v amb For blind speed, γ k Let v be the Doppler ambiguity number corresponding to the k-th target. k,dect =mod(v k ,v amb )satisfy The pulse repetition frequency is PRF = f PRF Then when the Doppler frequency range is At that time, blind speed is calculated as

[0057] Pulse compression is performed using matched filtering, followed by a fast-time FFT to obtain data coupling the slow-time velocity and the fast-time frequency, expressed by the formula:

[0058]

[0059] Among them, t m = mT, where f is the fast time frequency, and we have:

[0060] For the detection speed term corresponding to the k-th target, This is the velocity fuzzy term corresponding to the k-th target.

[0061] Step 2: Resample in the slow time dimension to obtain data after decoupling velocity and distance:

[0062] The sinc interpolation method is used to resample the slow time dimension, and the velocity fuzzy term is approximated as H. key (f,τ m )≈H(f,τ m This allows us to obtain data after decoupling the slow-time dimension velocity from the fast-time frequency, expressed by the formula:

[0063]

[0064] Where m is the original sampling point, m' is the sampling point after sinc interpolation, and τ m The resampling time after interpolation is slow. This is the resampling detection speed term corresponding to the k-th target. This is the resampling rate fuzzy term corresponding to the k-th target.

[0065] Step 3: Construct a matched filter for the velocity ambiguity number of the decoupled data, perform velocity ambiguity number search, and perform velocity ambiguity compensation on the decoupled data to obtain corrected data:

[0066] The formula for resampling to address velocity ambiguity is then simplified to represent the relationship between the true velocity and the blind velocity, expressed as:

[0067]

[0068] Among them, P k,real (f,τ m ) represents the resampled true velocity term corresponding to the k-th target.

[0069] For the velocity ambiguity number range of γ min <γ<γ max Constructing a matched filter Matched filtering is applied to the resampled decoupled data, expressed by the formula:

[0070]

[0071] For matched filter data R match Performing Inverse Fast Fourier Transform (IFFT) in the fast time dimension and Fast Fourier Transform (FFT) in the slow time dimension yields the distance-Doppler information matrix R corresponding to the velocity ambiguity number γ compensation. γ (n,m);

[0072] Take R γ The maximum value of the (n,m) modulus is selected, and the protected units near the maximum value are removed. The training units are then selected to construct the dataset corresponding to the velocity fuzzy number γ.

[0073] The estimated level of interference in the dataset is calculated using the following formula:

[0074]

[0075] Where, N r Indicates the number of training units. The unit number in the training unit set Γ(i,j) is represented by i and j, where i and j represent the coordinates of the training unit.

[0076] The compensation information for the corresponding velocity ambiguity number γ is calculated using the interference level estimate, and is expressed by the formula:

[0077]

[0078] Within the range of velocity ambiguity numbers, the corresponding compensation information is calculated sequentially, and the γ corresponding to the maximum value of the compensation information is taken as the estimated value of the velocity ambiguity number, expressed by the formula:

[0079]

[0080] And the corresponding velocity fuzzy number obtained after compensation is Subsequent corrected data Expressed as a formula:

[0081]

[0082] Step 4: Perform linear transformation on the corrected data to obtain the target data: In this embodiment, the linear transformation includes performing an inverse fast Fourier transform (IFFT) on the corrected data in the fast time dimension and a fast Fourier transform (FFT) in the slow time dimension to obtain the target data R(n,m) in the distance-Doppler dimension, which can be expressed by the formula:

[0083]

[0084] Where λ = c / f c .

[0085] Step 5: Extract the first data in the distance dimension and the second data in the velocity dimension from the target data, and calculate the square of the amplitude modulus of the first and second data. Use the highest value of the square of the amplitude modulus as the input of the pre-constructed two-dimensional constant false alarm detector for detection. If the target is detected as a valid target, the distance, velocity and amplitude of the target in the input target data are added to the set of valid target parameters; otherwise, the corresponding target data is removed.

[0086] In this embodiment, the determination of whether a target exists at the detection unit is expressed by the following formula:

[0087]

[0088] in, This represents the threshold used to determine a valid target. If the condition is met, it is determined that there is a valid target at the detection unit, and the next step of target parameter estimation can be performed; otherwise, the algorithm stops.

[0089] The echo signal can be represented in matrix form as follows:

[0090]

[0091] in,

[0092]

[0093] a M (ω)=[1,e jω ,…,e j(M-1)ω ] T and

[0094] Let y = vec(Y), w = vec(W). After flattening the matrix into a vector, we get:

[0095]

[0096] Then, the maximum likelihood estimation of the amplitude x, distance r, and velocity v corresponding to the first detection unit. This can be obtained by minimizing the energy of the residuals, i.e., a least squares problem:

[0097]

[0098] Where argmin(·) represents the optimal... After finding the minimum value of the entire expression, return the corresponding variable value.

[0099] In this embodiment, The horizontal and vertical coordinates corresponding to the maximum value of the magnitude of the distance-Doppler information matrix R(n,m) and The one-to-one correspondence is obtained, which can be expressed by the formula:

[0100]

[0101] Since the objective function is a quadratic function of the magnitude x, when r and v have been estimated, the optimal x is:

[0102]

[0103] Where N p =F s T p M is the number of pulses.

[0104] Finally, the distance, Doppler frequency, and amplitude value of the first detection unit that has completed detection are added to the set of valid targets.

[0105] Step 6: Repeat steps 1-5 until the termination condition of the two-dimensional constant false alarm rate detector is met, so as to output the final set of valid target parameters.

[0106] Assume the set of valid targets after detection is:

[0107]

[0108] The residual of the signal is then:

[0109]

[0110] The residual signal y r The data is fed into the data correction module, and the corrected data is then fed into the constant false alarm rate (CFAR) detection module to determine if any additional targets exist. If none exist, the algorithm stops. Otherwise, target parameters are estimated.

[0111] The distance-velocity information of all targets in the valid target set is output until all target information in the echo signal has been detected or the maximum number of iterations of the two-dimensional constant false alarm detector has been reached.

[0112] This embodiment also provides a high-speed target detection device, such as... Figure 2 As shown, it includes an echo data processing unit 510, a resampling unit 520, a velocity ambiguity correction unit 530, a target data acquisition unit 540, a parameter estimation unit 550, and a target output unit 560.

[0113] The echo data processing unit 510 is used to acquire the echo signal of the pulse Doppler radar and perform linear processing to obtain slow time dimension phase-coupled data.

[0114] The resampling unit 520 is used to resample the phase-coupled data in the slow time dimension to obtain the decoupled data;

[0115] The velocity ambiguity correction unit 530 is used to construct a matched filter for velocity ambiguity numbers on the decoupled data, search for velocity ambiguity numbers, and perform velocity ambiguity compensation on the decoupled data to obtain corrected data.

[0116] The target data acquisition unit 540 is used to perform a linear transformation on the corrected data to obtain the target data;

[0117] The parameter estimation unit 550 is used to calculate the square of the amplitude magnitude of the two-dimensional data in the distance and velocity dimensions of the target data, and takes the point with the highest value of the square of the amplitude magnitude as the first detection unit, inputs it into the pre-constructed two-dimensional constant false alarm detector for detection, and adds the distance, velocity and amplitude values ​​corresponding to the first detection unit that is detected as a valid target to the set of valid target parameters;

[0118] The target output unit 560 is used to remove the first result from the echo signal and repeat the echo data processing unit, resampling unit, velocity ambiguity correction unit, target data acquisition unit, and parameter estimation unit until the judgment termination condition of the two-dimensional constant false alarm detector is met, and output the final set of valid targets to obtain the distance and velocity information of all targets in the echo signal.

[0119] To better illustrate the technical effects of the method provided in this embodiment, the following test scenario was set: Pulse radar parameter settings in the experiment: Carrier frequency f c =3GHz; pulse width T p = 2μs; pulse repetition interval T = 100μs; bandwidth B = 3MHz; chirp rate μ = B / T p Blind speed v amb =500m / s; sampling frequency F s=7MHz; pulse number =1024, corresponding to a distance resolution of 21.4m and a speed resolution of 0.49m / s.

[0120] Target parameter settings: Target 1 velocity v1 = 2400 m / s, distance r1 = 6006 m; Target 2 velocity v2 = 1450 m / s, distance r2 = 5576 m; Target 3 velocity v3 = -1880 m / s, distance r3 = 4719 m. The velocity ambiguities for the three targets are γ1 = 5, γ2 = 3, and γ3 = -2, respectively; the cumulative signal-to-noise ratio (SNR) for all targets is 18 dB; the false alarm probability P0 is... fa =10 -8 The target spans 11.46, 6.93 and -8.98 distance units respectively (where "-" indicates the opposite direction).

[0121] The high-speed target detection method for pulse Doppler radar provided in this embodiment is used to detect various targets in the test scenario:

[0122] like Figure 3 The image shows the detection results for target one in the test scenario. Figure 3 The white box in the middle represents the reference range of the actual target parameters for Target 1.

[0123] like Figure 4 The image shows the detection results for target two in the test scenario. Figure 4 The white box in the middle represents the reference range of the actual target parameters for Target 2.

[0124] like Figure 5 The image shows the detection results for target three in the test scenario. Figure 4 The white box in the middle represents the reference range of the actual target parameters for Target 2.

[0125] like Figure 6 As shown, the velocity ambiguity number detection results for three targets are presented. Based on the results, it can be calculated that the distance error is less than one distance resolution unit, the velocity error is less than one distance resolution unit, and the velocity ambiguity number estimation accuracy is high.

[0126] Based on the same test scenario, the traditional pulse compression method based on matched filtering is used for detection:

[0127] like Figure 7 The image shows the detection results for target one in the test scenario. Figure 8 The image shows the detection results for target two in the test scenario. Figure 9 The image shows the detection results for target three in the test scenario. The target energy was too dispersed due to the matched filtering method's inability to correct distance migration, resulting in its failure to be identified. Figure 7 Target 1 and Figure 8 The second objective is to target... Figure 9 In the case of Target 3, the number of targets for Target 3 was overestimated due to the strong dispersion of energy.

[0128] In summary, traditional pulse compression methods cannot handle velocity ambiguity, thus limiting the velocity estimation range to -250m / s to 250m / s, making it impossible to estimate the velocity of high-speed targets.

[0129] The high-speed target detection method provided in this embodiment can meet the requirements for distance and velocity estimation of targets under high-speed conditions.

Claims

1. A method for high-speed target detection against pulse Doppler radar, characterized in that, The method comprises the following steps: Step 1, obtaining echo signals transmitted by a pulse Doppler radar, and performing linear processing on the echo signals to obtain initial data coupled by velocity and distance of each target; Step 2, resample the initial data in the slow time dimension to obtain resampled data after decoupling velocity and distance, the expression of the resampling is as follows: ; wherein, is the original sampling point, is the sinc-interpolated sampling point, is the fast time frequency, is the carrier frequency, is the number of pulses; step 3, compensate the resampled data by a pre-constructed velocity blurring number matched filter to obtain corresponding correction data, the expression of the correction data is as follows: ; wherein, represents the velocity blurring number used to compensate the resampled data, is the virtual time, represents the velocity blurring number matched filter; step 4, perform linear transformation processing on the correction data to obtain target data; Step 5, extracting first data in the distance dimension and second data in the velocity dimension from the target data, and calculating amplitude modulus squares of the first data and the second data, taking a highest value point of the amplitude modulus squares as an input of a pre-constructed two-dimensional constant false alarm detector for detection; If the detection is an effective target, distance, velocity and amplitude of the target in the input target data are added to an effective target parameter set; Otherwise, the corresponding target data is removed; Step 6, repeating steps 1-5 until a judgment termination condition of the two-dimensional constant false alarm detector is met, to output a final effective target parameter set.

2. The method for high-speed target detection against pulse Doppler radar according to claim 1, characterized in that, The pulse Doppler radar is a shore-based radar, an airborne radar or a shipborne radar.

3. The method for high-speed target detection against pulse Doppler radar according to claim 1, characterized in that, In step 1, a specific process of the linear processing is as follows: A matched filter method is used to perform pulse compression on the echo signals, and fast time dimension FFT is performed on signal data obtained by the pulse compression to obtain initial data coupled by slow time dimension phase.

4. The method for high-speed target detection against pulse Doppler radar according to claim 1, characterized in that, In step 2, resampling is done by a sinc interpolation method in the slow time dimension under the condition that where, is the virtual time, is the original time.

5. The method for high-speed target detection against pulse Doppler radar according to claim 1, characterized in that, In step 4, a process of the linear transformation processing is as follows: Inverse fast Fourier transform is performed on the modified data in the fast time dimension, and fast Fourier transform is performed on the modified data in the slow time dimension to obtain target data in the distance-Doppler dimension.

6. The method for high-speed target detection against pulse Doppler radar according to claim 1, characterized in that, In step 5, a detection process of the two-dimensional constant false alarm detector is as follows: All training units are removed around the input amplitude modulus square maximum value; An average value of all data except the amplitude modulus square maximum value is calculated to obtain a noise floor value; Based on the noise floor value, a threshold value of an effective target is determined, and a ratio of the amplitude modulus square maximum value to the noise floor value is compared with the threshold value: If the ratio is greater than the threshold value, it is considered that there is an effective target, otherwise, it is considered that there is a non-effective target.

7. The method for high-speed target detection against pulse Doppler radar according to claim 1, characterized in that, The judgment termination condition comprises: a, a repetition number has reached a maximum iteration number preset by the constant false alarm detector; b, the constant false alarm detector judges that there is no target, that is, a detection statistic corresponding to the maximum value point does not exceed a corresponding threshold value.

8. A high-speed target detection apparatus characterized by comprising: A device for executing steps of the high-speed target detection method for the pulse Doppler radar according to any one of claims 1-7.

Images