A method for angle estimation of high-speed weak targets in space-time joint radar

Abstract

The invention discloses a space-time joint radar high-speed weak target angle estimation method. Firstly, beamforming processing is performed on array element-level received data to make the radar beam point to the direction of the target to be distinguished, and then a Stepped-GRFT algorithm is used to perform long-term coherent processing on the echo data after beamforming to complete target detection, obtain a radial motion parameter estimation vector, select a distance unit where the target to be distinguished is located, use the data corresponding to the distance unit as the target array element-level data, and perform Doppler compensation; construct a space-time joint matching filter to obtain an array element-level space-time joint coherent accumulation result, search in the parameter space, and find the maximum angle-dimensional motion parameter set of the target to be distinguished as the estimation result; the algorithm performs long-term coherent accumulation to ensure the radial motion parameter estimation accuracy and the Doppler compensation accuracy; the space-time joint matching filter performs filtering processing on the coherent accumulation to ensure the estimation accuracy of the angle motion parameter.

Description

Technical Field

[0001] The present invention relates to the technical field of radar signal processing, and in particular to a space-time joint radar high-speed weak target angle estimation method. Background Art

[0002] Estimating the motion parameters of high-speed weak targets is an important task in the field of radar detection. For radar observable targets, the so-called high-speed weak targets refer to targets with speeds of hundreds to thousands of meters per second and radar cross-sectional areas of less than 0.01 square meters. At this speed, the target will have obvious cross-range unit phenomenon within the coherent accumulation time. Considering this problem from the perspective of radar signal processing, for high-speed weak targets, it is first necessary to obtain sufficient signal-to-noise ratio gain through long-term accumulation methods such as GRFT algorithm and Stepped-GRFT algorithm to achieve target detection and radial motion parameter estimation, and then select a suitable angle measurement algorithm to complete the target angle estimation. However, under the condition of long-term accumulation (in long-term estimation, the coherent processing cycle is generally set to seconds), the angle change of the target will not be ignored. It is of great practical significance to study the target angle motion model and parameter estimation method.

[0003] Traditional radar target angle measurement methods mainly include equal signal methods based on sum and difference beams and direction of arrival (DOA) estimation methods based on spatial spectrum estimation. Equal signal methods such as amplitude or phase angle measurement methods are simple in principle and easy to implement, but the angle resolution is relatively low; while DOA methods suitable for array radars such as the multiple signal classification (MUSIC) method and the rotation factor invariant (ESPRIT) rule can achieve angle super-resolution. However, the above traditional angle measurement methods are all studied for single pulse radar target angle measurement, which have high requirements on the target signal-to-noise ratio and cannot be applied to scenarios where the target angle changes.

[0004] Therefore, studying a method for estimating the angle parameters of radar high-speed weak targets under long-term accumulation conditions has important practical significance and application value. Summary of the invention

[0005] In view of this, the present invention provides a space-time joint radar high-speed weak target angle estimation method, which can solve the problem of failure of traditional single-pulse angle measurement methods caused by changes in target angle under long-term accumulation conditions, so as to obtain accurate angle information of the target and achieve precise target positioning.

[0006] In order to achieve the above-mentioned object of the invention, the technical solution of the present invention is:

[0007] A space-time joint radar high-speed weak target angle estimation method uses an array radar with L array elements to perform long-term angle estimation on a high-speed weak target. The estimation time includes M preset coherent processing cycles T, and each coherent processing cycle T contains P pulses. The specific steps of the estimation method include:

[0008] Step 1: Perform beamforming processing on the radar's element-level received data to point the radar beam in the direction of the target to be identified.

[0009] Step 2: Use the Stepped-GRFT algorithm to perform time-coherent processing on the echo data after beamforming to complete target detection and obtain the radial motion parameter estimation vector.

[0010] Step 3: Obtain the distance unit where the target to be resolved is located in each pulse within the estimation time according to the radial motion parameter estimation vector, take out the corresponding distance unit data from all array data and combine them according to the pulse sequence to obtain the target array element-level data, that is, the set of array element-level data of the distance unit where the target to be resolved is located.

[0011] Step 4: Perform Doppler compensation on the target array element-level data to obtain Doppler compensated array element-level data.

[0012] Step 5: construct a space-time joint matching filter based on the target element-level data obtained in step 3; use the space-time joint matching filter to perform angle phase compensation on the element-level data after Doppler compensation, and sum them up to obtain the element-level space-time joint coherent accumulation result.

[0013] Step 6: Search in the parameter space to find the set of target angle-dimensional motion parameters that maximizes the result of the element-level space-time joint coherent accumulation, which is the estimated result of the target angle-dimensional motion parameters.

[0014] Furthermore, the radial motion parameter estimation vector is The formula is:

[0015]

[0016] Wherein, T is the preset coherent processing cycle, and T contains P pulses in one coherent processing cycle; K is the number of radial motion parameters, α0 is the initial radial distance of the target to be distinguished, α1 is the first-order derivative of the initial radial distance of the target to be distinguished, α2 is the second-order derivative of the initial radial distance of the target to be distinguished, and α K-1 is the K-1 derivative of the initial radial distance of the target to be resolved.

[0017] Furthermore, the target element-level data is The formula is:

[0018]

[0019] in, is the spatial guidance vector, d is the interval between adjacent array elements, s0 is the target complex scattering coefficient, θ is the target angle phase, R(p) is the Doppler phase term, is a zero-mean Gaussian white noise sequence, is the target angle phase set; p is the pulse number, l is the array element number; p=0,1,2,...,P-1; l=0,1,2,...,L-1; is the angle-dimensional motion parameter set of the target to be distinguished.

[0020] Furthermore, the array element level data is The formula is:

[0021]

[0022] Among them, j is an imaginary number, c is the speed of light, and f n is the carrier frequency of each pulse in the frame, n=0,1,2,...,N-1 is the number of pulses of the carrier frequency; is the instantaneous radial distance of the target to be resolved.

[0023] Furthermore, the space-time joint matched filter The formula is:

[0024]

[0025] in, For the kth w The slow time corresponding to each pulse, ω T0 k is the first element of the target angle motion parameter set to be resolved; w is the sequence number of the motion parameter, k w =0,1,2...,K w -1, K w is the number of motion parameters; is the kth angle-dimensional motion parameter set of the target to be distinguished w +1 element.

[0026] Furthermore, the setting principle of the coherent processing period T is: within a coherent processing period T, the target to be resolved is always located in the main lobe of the beam;

[0027] The width of the main lobe BW is:

[0028]

[0029] Where λ is the radar wavelength and d is the interval between adjacent array elements.

[0030] Furthermore, in step 6, the parameter space is searched, and the parameter search range of the target initial angle θ0 and the target angle dimension initial parameter ω is set to:

[0031]

[0032] Furthermore, the formula for the pulse number p is expressed as:

[0033] p=mN+n,m=0,1,…,M-1 (7)

[0034] Wherein, m is the number of coherent processing cycles T.

[0035] Furthermore, the carrier frequency f of each pulse in the frame n The formula is:

[0036] f n =f0+d(n)Δf,n=0,1,…,N-1 (8)

[0037] Among them, f0 is the initial carrier frequency, Δf is the frequency step interval, and d(n)∈[0,1,…,N-1] is the carrier frequency coding sequence.

[0038] Furthermore, the instantaneous radial distance of the target to be resolved The formula is:

[0039]

[0040] Among them, α Tk is the initial radial distance Tk derivative of the target to be resolved, is the slow time corresponding to the kth pulse.

[0041] Beneficial effects:

[0042] 1. The present invention proposes a space-time joint radar high-speed weak target angle estimation method, which solves the problem of failure of traditional single-pulse angle measurement methods caused by changes in target angles under long-term accumulation conditions, so as to obtain accurate angle information of the target and achieve precise positioning of the target. The method first performs beamforming processing on the array-level received data to point the radar beam to the direction of the target to be distinguished, and then uses the Stepped-GRFT algorithm to perform long-term coherent processing on the array-level echo data after beamforming to complete target detection and obtain radial motion parameter estimation vectors; according to the radial motion parameter estimation vectors, the distance unit where the target to be distinguished is located is selected, and the data corresponding to the distance unit is used as the target array-level data, and Doppler compensation is performed; a space-time joint matching filter is constructed to obtain the array-level space-time joint coherent accumulation result, and the parameter space is searched to find the maximum angle-dimensional motion parameter set of the target to be distinguished, which is the estimation result of the target angle-dimensional motion parameter.

[0043] 2. The present invention uses the Stepped-GRFT algorithm to perform coherent processing on high-speed weak targets to obtain radial motion parameter estimation vectors. The algorithm can ensure the accuracy of radial motion parameter estimation and Doppler compensation during a long-term accumulation process. On the basis of high-precision radial parameters, the present invention further estimates angle parameters, thereby ensuring that the accuracy of angle parameters is higher.

[0044] 3. The present invention adopts a space-time joint matching filter, and on the basis of obtaining high-precision radial parameters, performs filtering processing on the coherent accumulation results to compensate for the angle phase. Further, on the basis of compensating the angle phase, the element-level space-time joint coherent accumulation results are obtained. A search is performed in the preset parameter space to further obtain an estimate of the angle parameter to achieve angle estimation.

[0045] 4. The present invention models the moving target and obtains an estimation method for angle parameters. That is, the method of the present invention takes into account the change in target angle caused by the movement, so a long-term estimation calculation can be selected while ensuring measurement accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 The figure is a flow chart of the method of the present invention.

[0047] Figure 2 Schematic diagram of two-dimensional uniform target motion.

[0048] Figure 3 It is the target motion path diagram.

[0049] Figure 4 The accuracy diagram of target initial angle estimation under different signal-to-noise ratios.

[0050] Figure 5 The target angular velocity estimation accuracy diagram under different signal-to-noise ratios. DETAILED DESCRIPTION

[0051] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0052] Based on the two-dimensional linear motion model, this paper proposes a space-time joint radar high-speed weak target angle estimation algorithm, which realizes accurate estimation of target angle dimension motion parameters by designing a space-time joint matched filter bank and parameter search. This method is applicable to a variety of different radar waveforms including frequency stepping signals and repetition frequency jitter signals.

[0053] To achieve the above object, the present invention is achieved through the following technical solutions:

[0054] Long-term angle estimation is performed on an array radar with L array elements. Assume that the accumulation time corresponding to one coherent processing interval (CPI, i.e., one coherent processing cycle) of the radar system is T. The estimation time includes M preset coherent processing cycles T. Each coherent processing cycle T contains P pulses. The carrier frequency of each pulse in the frame is:

[0055] f n =f0+d(n)Δf,n=0,1,…,N-1 (8)

[0056] Where f0 is the initial carrier frequency, Δf is the frequency step interval, d(n)∈[0,1,…,N-1] is the carrier frequency coding sequence, n is the number of carrier frequency pulses, and N is the total number of pulses within a period T.

[0057] In the embodiment of the present invention, the number of coherently accumulated pulses is P=MN, and the formula for the pulse number p is expressed as:

[0058] p=mN+n,m=0,1,…,M-1 (7)

[0059] Among them, m is the number of cycles in the coherent accumulation practice, and M is the total number of cycles in the coherent accumulation practice.

[0060] The pulse repetition time (PRT) corresponding to the pth pulse is T r (p), whose value jitters between different pulses, and p is the pulse number. Therefore, the slow time corresponding to each pulse in a CPI can be expressed as

[0061]

[0062] Considering a single target scenario, assuming that the radial motion parameters of the target are expressed as Represented as follows, where α0 is the initial radial distance of the target to be resolved, α1 is the first-order derivative of the initial radial distance of the target to be resolved, α2 is the second-order derivative of the initial radial distance of the target to be resolved, and α K-1 is the K-1 derivative of the initial radial distance of the target to be resolved. Then the instantaneous radial distance of the target is:

[0063]

[0064] Among them, α Tk is the initial radial distance Tk derivative of the target to be resolved, is the slow time corresponding to the kth pulse.

[0065] Based on the linear radial motion model of equation (9), assuming that the target angle dimension also moves linearly within a certain observation time, the target angle phase set can be expressed as:

[0066]

[0067] in, For the kth w The slow time corresponding to each pulse, ω T0 k is the first element of the target angle motion parameter set to be resolved; w is the sequence number of the motion parameter, k w =0,1,2...,K w -1, K w is the number of motion parameters; is the kth angle-dimensional motion parameter set of the target to be distinguished w +1 element.

[0068] The target motion constrained by equations (9) and (10) is defined as a two-dimensional linear motion model. When only the radial velocity and angular velocity of the target are considered, the model will degenerate into a two-dimensional uniform motion model. Figure 2 Schematic diagram of the motion path of a two-dimensional uniformly moving target with a single scattering point.

[0069] Assume that the radar array is a one-dimensional uniform linear array consisting of L array elements. Figure 3 As shown in the figure, the target moves at a uniform speed in both radial and angular dimensions, with radial velocity v and angular velocity ω. According to the target radial motion parameter estimation results obtained by the Stepped-GRFT algorithm, the data of the distance unit where the target is located can be extracted from the P×L echoes at the array element level, and the array element level target data is obtained as follows:

[0070]

[0071] in, is the target angle dimension motion parameter set, is the spatial guidance vector, d is the interval between adjacent array elements, s0 is the target complex scattering coefficient, R(p) is the Doppler phase term, is a zero-mean Gaussian white noise sequence, is the target angle phase set, and θ is the target angle phase.

[0072]

[0073] For simplicity, assuming that it does not change with the transmission frequency, the Doppler phase term R(p) determined by the target radial motion parameter is:

[0074]

[0075] like Figure 1 As shown, the invention provides a space-time joint radar high-speed weak target angle estimation method, the specific steps include:

[0076] Step 1: Perform beamforming processing on the array-level received data to make the radar beam point to the direction of the target to be identified.

[0077] Step 2: Use the Stepped-GRFT algorithm to perform long-term coherent processing on the element-level echo data after beamforming to complete target detection and obtain the radial motion parameter estimation vector The Stepped-GRFT algorithm can ensure the estimation accuracy of radial parameters in a long-term process and reduce the error caused by angle changes. Based on the high-precision radial parameters, the present invention further estimates the angular parameters, thereby ensuring a higher accuracy of the angular parameters.

[0078]

[0079] Among them, T is the coherent processing period, K is the number of radial motion parameters, α0 is the initial radial distance of the target to be resolved, α1 is the first-order derivative of the initial radial distance of the target to be resolved, α2 is the second-order derivative of the initial radial distance of the target to be resolved, and α K-1 is the K-1 derivative of the initial radial distance of the target to be resolved.

[0080] Step 3: Estimate vector based on radial motion parameters Get target element-level data That is, the set of array element-level data of the distance unit where the target to be distinguished is located. The specific method is: according to the radial motion parameter estimation vector In each pulse, the distance unit where the target to be distinguished is located is combined with the data of the corresponding distance unit according to the pulse sequence to obtain the target array element level data. Its expression can be written as:

[0081]

[0082] in, is the spatial guidance vector, d is the interval between adjacent array elements, s0 is the target complex scattering coefficient, θ is the target angle phase, R(p) is the Doppler phase term, is a zero-mean Gaussian white noise sequence.

[0083] p is the pulse number, l is the array element number; p = 0, 1, 2, ..., P-1; l = 0, 1, 2, ..., L-1; P is the number of pulses, L is the number of array elements.

[0084] Step 4: Target array element level data Perform Doppler compensation to obtain the element-level data after Doppler compensation

[0085]

[0086] Among them, j is an imaginary number, c is the speed of light, and f n is the carrier frequency of each pulse in the frame, n=0,1,2,...,N-1 is the number of pulses of the carrier frequency, and N is the total number of pulses in one cycle; is the instantaneous radial distance of the target to be resolved.

[0087] Step 5: Target element-level data obtained from step 3 Constructing a space-time joint matched filter The specific method is: according to the target array element level data And formula (2), we get the spatial steering vector Its formula is shown in formula (11). By taking the conjugate of formula (11), we can get the result of formula (4).

[0088] After the construction is completed, the space-time joint matched filter bank is used right Perform angle phase compensation and sum to obtain the element-level space-time joint coherent accumulation result On the basis of obtaining high-precision radial parameters, the coherent accumulation results are filtered using a space-time joint matched filter group to compensate for the angle phase and improve the measurement accuracy.

[0089]

[0090] in, is the angle-dimensional motion parameter set of the target to be distinguished.

[0091]

[0092] in, For the kth w The slow time corresponding to each pulse, ω T0 k is the first element of the target angle motion parameter set to be resolved; w is the sequence number of the motion parameter, k w =0,1,2...,K w -1, K w is the number of motion parameters; is the kth angle-dimensional motion parameter set of the target to be distinguished w +1 element.

[0093] Step 6: Search in the parameter space to find The largest That is the estimated result of the target angle dimension motion parameter. Step six is ​​expressed by the formula:

[0094]

[0095] In the embodiment of the present invention, within a coherent processing period T, the target to be resolved is always located in the main lobe of the beam.

[0096] In the embodiment of the present invention, the width BW of the main lobe is:

[0097]

[0098] Where λ is the radar wavelength and d is the interval between adjacent array elements.

[0099] In the embodiment of the present invention, when designing the angle-dimensional motion parameter search space, the parameter search range of the target initial angle θ0 and the target angle-dimensional initial parameter ω is:

[0100]

[0101] In the embodiment of the present invention, the formula for the instantaneous radial distance of the target to be resolved is expressed as:

[0102]

[0103] Among them, α Tk is the initial radial distance Tk-order derivative of the target to be resolved, is the slow time corresponding to the kth pulse.

[0104] A comparative simulation example is given below for the present invention, and its implementation process is analyzed and explained.

[0105] 1. Comparison of target initial angle estimation accuracy under different signal-to-noise ratios:

[0106] The parameters are set as follows: the number of array elements is set to 20, the array element spacing is 0.5λ=0.1m, the target initial angle is set to -0.03rad (-1.72°), the target angular velocity is 0.12rad / s (6.88° / s), the target initial radial distance is 33km, the target radial velocity is 705m / s, and the array element level signal-to-noise ratio after pulse compression is -30dB~10dB. During the search, the radar beam main lobe width is calculated to be 5.07°, so the target initial angle search range is -2.55°~2.55°, and the angular velocity search range is -10° / s~10° / s. 500 Monte Carlo simulations are performed to obtain the statistical results of the target initial angle estimation accuracy, as shown in the figure. Figure 4 shown.

[0107] 2. Comparison of target angular velocity estimation accuracy under different signal-to-noise ratios:

[0108] The simulation settings are the same as above. Under different signal-to-noise ratio conditions, 500 Monte Carlo simulation experiments were performed, and the target signal-to-noise ratio also changed linearly between -30dB and 10dB. The simulation results are shown in Figure 5 shown.

[0109] The results of 1 and 2 show that the target initial angle and angular velocity estimation results converge to the true value when the signal-to-noise ratio is -10dB and -12dB, respectively, completing the accurate estimation of the target angle-dimensional motion parameters.

[0110] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A space-time joint radar high-speed weak target angle estimation method, characterized in that: An array radar with L array elements is used to perform long-term angle estimation on high-speed weak targets. The estimation time includes M preset coherent processing cycles T, and each coherent processing cycle T contains P pulses. The specific steps of the estimation method include: Step 1: Perform beamforming processing on the radar element-level received data to make the radar beam point to the direction of the target to be identified; Step 2: Use the Stepped-GRFT algorithm to perform time coherent processing on the echo data after beamforming, complete target detection, and obtain the radial motion parameter estimation vector; Step 3: Obtain the distance unit where the target to be resolved is located in each pulse within the estimation time according to the radial motion parameter estimation vector, take out the corresponding distance unit data from all array data and combine them according to the pulse sequence to obtain the target array element level data, that is, the set of array element level data of the distance unit where the target to be resolved is located; Step 4: performing Doppler compensation on the target array element-level data to obtain Doppler-compensated array element-level data; Step 5: construct a space-time joint matched filter based on the target element-level data obtained in step 3; use the space-time joint matched filter to perform angle phase compensation on the element-level data after Doppler compensation, and sum them to obtain the element-level space-time joint coherent accumulation result; Step 6: Search in the parameter space to find the set of target angle-dimensional motion parameters that maximizes the result of the element-level space-time joint coherent accumulation, which is the estimated result of the target angle-dimensional motion parameters.

2. The method according to claim 1, characterized in that The radial motion parameter estimation vector is The formula is: Wherein, T is the preset coherent processing cycle, and T contains P pulses in one coherent processing cycle; K is the number of radial motion parameters, α0 is the initial radial distance of the target to be distinguished, α1 is the first-order derivative of the initial radial distance of the target to be distinguished, α2 is the second-order derivative of the initial radial distance of the target to be distinguished, and α K-1 is the K-1 derivative of the initial radial distance of the target to be resolved.

3. The method according to claim 2, characterized in that The target element-level data is ∧x(p,l), which is expressed as: in, is the spatial guidance vector, s0 is the target complex scattering coefficient, θ is the target angle phase, R(p) is the Doppler phase term, is a zero-mean Gaussian white noise sequence, is the target angle phase set; p is the pulse number, l is the array element number; p=0,1,2,...,P-1; l=0,1,2,...,L-1; is the angle-dimensional motion parameter set of the target to be distinguished, and j is an imaginary number.

4. The method according to claim 3, characterized in that The array element level data is The formula is: Among them, j is an imaginary number, c is the speed of light, and f n is the carrier frequency of each pulse in the frame, n=0,1,2,...,N-1 is the number of pulses of the carrier frequency; is the instantaneous radial distance of the target to be resolved.

5. The method according to claim 4, characterized in that The space-time joint matched filter The formula is: in, For the kth w The slow time corresponding to each pulse, ω T0 k is the first element of the target angle motion parameter set to be resolved; w is the sequence number of the motion parameter, k w =0,1,2...,K w -1, K w is the number of motion parameters; is the kth set of angle-dimensional motion parameters of the target to be distinguished w +1 element, d is the interval between adjacent array elements.

6. The method according to any one of claims 1 to 5, characterized in that: The setting principle of the coherent processing period T is: within a coherent processing period T, the target to be distinguished is always located in the main lobe of the beam; The width BW of the main lobe is: Where λ is the radar wavelength and d is the interval between adjacent array elements.

7. The method according to claim 6, characterized in that In step 6, the parameter space is searched, and the parameter search range of the target initial angle θ0 and the target angle dimension initial parameter ω is set to:

8. The method according to claim 4, characterized in that The formula for the pulse number p is expressed as: p=mN+n, m=0,1,…,M-1 (7) Wherein, m is the number of coherent processing cycles T.

9. The method according to claim 4, characterized in that The carrier frequency f of each pulse in the frame n The formula is: f n =f0+d(n)Δf, n=0,1,…,N-1 (8) Where f0 is the initial carrier frequency, Δf is the frequency step interval, and d(n)∈[0,1,…,N-1] is the carrier frequency coding sequence.

10. The method according to claim 4, characterized in that The instantaneous radial distance of the target to be resolved The formula is: Among them, α Tk is the initial radial distance Tk-order derivative of the target to be resolved, is the slow time corresponding to the kth pulse.

Images