Method for estimating elevation angle of low-altitude target

Abstract

The invention relates to the field of radar signal processing, in particular to a method for estimating the elevation angle of a low-altitude target. According to the method, different terrain conditions are comprehensively considered, the RML algorithm and the multi-frequency beam scanning algorithm are fused for processing, in the flat terrain environment, the RML algorithm is adopted to be combined with signal space-time information, distance information and a prior reflection coefficient, a composite steering vector under the multi-path condition is generated to estimate the elevation angle of the target, and the elevation angle of the target is estimated. The estimation mode utilizes various prior information and has the characteristics of low calculation amount and high angle measurement precision, and aiming at the fluctuating terrain environment, a multi-frequency beam scanning algorithm is adopted, so that the angle measurement precision interference caused by relatively strong ground diffuse reflection components can be effectively avoided.

Description

Technical Field

[0001] The invention relates to the field of radar signal processing, and in particular to a method for estimating the elevation angle of a low-altitude target. Background Art

[0002] At present, in the field of radar target tracking, the detection and tracking of low-altitude targets has always been a challenging research direction. When tracking low-angle targets, the signals received by the mobile phased array radar include not only direct wave signals, but also multipath signals such as mirror reflection and diffuse reflection, as well as background noise. The presence of multipath signals will cause the radar's vertical lobe to split and warp, resulting in a sharp decline in the radar's target direction finding and tracking performance.

[0003] In order to detect long-distance low-altitude targets, we can make full use of the characteristics of phased array radar (multiple degrees of freedom, multi-target tracking, long effective range, high data rate, high reliability, etc.), adopt the new theories and methods developed in recent years for array signals, correctly separate and estimate the angles of each target, and improve the accuracy of elevation angle measurement of phased array radar during low-angle tracking.

[0004] Array super-resolution technology is an important means to improve the performance of radar target elevation angle estimation, among which the characteristic subspace algorithm (such as MUSIC algorithm) and maximum likelihood (ML) algorithm are the representatives. The former has relatively high requirements on the number of snapshots, so actual radar systems often do not use this method. Although the maximum likelihood algorithm often requires multi-dimensional search and has a relatively large amount of calculation, it does not require high radar pulse resources and is suitable for coherent source situations. Even in the case of a single snapshot, the estimated value of the target parameter can be obtained. Therefore, in actual engineering applications, the maximum likelihood algorithm is mainly discussed, mainly including: (1) spatiotemporal cascade ML algorithm; (2) improved spatiotemporal cascade ML algorithm; (3) exact maximum likelihood (RML) algorithm.

[0005] Method (1) utilizes the spatial and temporal information of the signal, so a two-dimensional search is required, which results in a large amount of computation. Method (2) is based on method (1) and utilizes the geometric relationship between the elevation angles of the direct signal and the multipath signal in the echo, thereby converting the two-dimensional search into a one-dimensional search to reduce the amount of computation. Method (3) utilizes the prior information of the reflection coefficient based on method (2), thereby further reducing the amount of computation and further improving the angle measurement accuracy.

[0006] The above ML algorithm is only applicable to the mirror reflection model with a flat reflecting surface. When the reflecting surface is undulating, the multipath signal contains both mirror reflection signals and diffuse reflection signals, and even the diffuse reflection component plays a major role. The signal reception model will change, and the performance of the above angle measurement algorithm will drop sharply. Summary of the invention

[0007] In view of the deficiencies in the prior art, the present invention provides a method for estimating the elevation angle of low-altitude targets. The method of the present invention comprehensively adopts the RML algorithm and the multi-frequency beam scanning algorithm in combination with different ground terrain conditions. The RML algorithm is adopted in flat terrain, and the composite guidance vector under multipath conditions is used to estimate the elevation angle of the target in combination with the signal time and space information, distance information and prior reflection coefficient. This estimation method utilizes a variety of prior information and has the characteristics of low computational complexity and high angle measurement accuracy. For undulating terrain environments, a multi-frequency beam scanning algorithm is adopted to avoid the interference of angle measurement accuracy caused by the strong ground diffuse reflection component. The present invention has the advantages of versatility, high computational efficiency, and easy implementation in engineering, and has been successfully verified for use in a certain model of radar.

[0008] The technical solution of the present invention is: a method for estimating the elevation angle of a low-altitude target, comprising the following steps:

[0009] Step 1: Obtain radar echo signals and perform coherent accumulation on the array signals.

[0010] The beneficial effect of this step is to superimpose multiple coherent echo signals, improve the signal-to-noise ratio, enhance the target detection performance, and improve the target parameter estimation accuracy.

[0011] Step 2: Separate the signal data according to the different terrain environments within the scanning direction, and select different algorithms for processing.

[0012] The beneficial effect of this step is to separate the signal models under different reception scenarios, and to achieve the best estimation accuracy by subsequently adopting an estimation algorithm suitable for the model.

[0013] Step 3: For flat terrain, use the RML algorithm for estimation. Perform conventional beamforming on the output data to point the spatial beam to the estimated target direction, perform signal detection on the data of each array element, perform spectrum analysis (FFT) on the detected target and detect the spectrum peak to obtain the target Doppler frequency estimation value of different array elements. Composition signal array vector The maximum likelihood estimation of the array composite steering vector w under multipath conditions is performed to obtain the likelihood function and the estimated value of the target elevation angle is obtained.

[0014] The beneficial effect of this step is that the precise maximum likelihood (RML) algorithm is used in flat terrain to target the dominant mirror reflection signal component in the environment, while reducing the computational complexity and improving the estimation accuracy. The RML algorithm is based on the multipath signal model under mirror conditions, and makes full use of three prior information including time and space dimensions, distance and reflection coefficient to form a composite guidance vector under multipath conditions to replace the conventional guidance vector in free space, reduce the search dimension, and separate each target through spectral analysis and use the maximum likelihood estimation method to estimate the target elevation angle. This algorithm can well adapt to the mirror reflection model of flat ground, accurately estimate the elevation angle of each target, and effectively reduce the amount of calculation.

[0015] Step 4: For the undulating terrain environment, a multi-frequency beam scanning algorithm is used for estimation. For the target echo data x1 and x2 obtained at the two working frequencies, the array steering vectors a1(θ) and a2(θ) are obtained respectively, and the likelihood function values ​​T1(θ) and T2(θ) are calculated and combined to obtain the target elevation angle estimate.

[0016] The beneficial effect of this step is to use a multi-frequency beam scanning algorithm in undulating terrain to target the dominant diffuse reflection signal component in the environment. In undulating terrain, the conventional mirror reflection signal model no longer matches the actual situation, and the RML algorithm is no longer applicable. This algorithm is based on the phenomenon that the main lobe peak position corresponding to the target elevation angle in the elevation spectrum at different operating frequencies remains basically unchanged, while the side lobe peak position shifts. The elevation spectra at different operating frequencies are added together to construct a multi-frequency elevation spectrum, so that the side lobe level will be reduced relative to the main lobe level, thereby improving the estimation accuracy of the elevation angle.

[0017] Step 5: Integrate all target elevation angle estimates, combine the elevation angle estimates for flat terrain and undulating terrain, and obtain the elevation angle measurement results within the entire radar scanning range.

[0018] The beneficial effect of this step is to comprehensively estimate the target elevation angle under different terrain conditions, which can not only significantly improve the radar angle measurement performance, but also make full use of diverse prior information, effectively reduce the algorithm calculation complexity and amount of calculation, and facilitate engineering implementation.

[0019] The beneficial effect of the present invention is that for the low-angle target elevation detection when the phased array radar detects low-altitude targets, an estimation algorithm that integrates terrain conditions is adopted to ensure the elevation angle measurement accuracy during low-angle tracking, improve the detection performance of the radar, and solve the problems of poor elevation angle estimation accuracy, large calculation dimension, and difficult engineering implementation in low-angle tracking of the traditional maximum likelihood algorithm.

[0020] This comprehensive low-angle target elevation angle estimation algorithm is a comprehensive use of different algorithms under different terrain conditions. In the flat terrain environment, the precise maximum likelihood estimation method is used, and the composite steering vector under multipath conditions is used to replace the conventional steering vector in free space, and then the maximum likelihood estimation method is used for estimation to reduce the computational dimension. In the undulating terrain environment, multi-frequency beams are used to superimpose the elevation angles at different working frequencies to construct a multi-frequency elevation spectrum, thereby improving the target elevation angle estimation accuracy. This algorithm enables the final detection algorithm to achieve better estimation results and can effectively reduce the amount of computation. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 The present invention is a flowchart of steps 1 to 4 in a specific embodiment of a method for low-altitude target detection and tracking based on a mobile phased array radar.

[0022] Figure 2 This is a curve diagram showing the RMSE of each maximum likelihood estimation algorithm changing with target height.

[0023] Figure 3 This is a curve of RMSE of the RML algorithm changing with target height for different antenna heights under the linear array.

[0024] Figure 4 This is the relationship between RMSE and SNR of the RML algorithm when there is an amplitude error in the reflection coefficient ρ under the linear array.

[0025] Figure 5 This is the relationship between RMSE and SNR of the RML algorithm when there is a phase error in the reflection coefficient ρ under the linear array.

[0026] Figure 6 The graphs of single-frequency beam (dotted line) and two multi-frequency combined beam scanning algorithms (dashed line and solid line) under linear array as the target elevation angle changes.

[0027] Figure 7 The curve of elevation angle error measured by RML algorithm under planar array and change with signal-to-noise ratio.

[0028] Figure 8 The curve of the elevation angle error measured by the RML algorithm under the planar array and the change of the target elevation angle.

[0029] Fig. 9 Curve diagram of single-frequency beam (dotted line) and two multi-frequency combination beam scanning algorithms (dashed line and solid line) under area array as the target elevation angle changes.

[0030] Fig.10 Comparison chart of the RML algorithm scanning curve under self-test data and the theoretical model.

[0031] Fig.11 The root mean square error of angle measurement varies with the signal-to-noise ratio under self-test data.

[0032] Fig.12 The root mean square error of angle measurement varies with target elevation angle under self-test data. DETAILED DESCRIPTION

[0033] The preferred implementation examples of the present invention are described in detail below in conjunction with the accompanying drawings so that the advantages and features of the present invention can be more easily understood by those skilled in the art, thereby making a clearer and more definite definition of the protection scope of the present invention.

[0034] The present invention provides a method for estimating the elevation angle of a low-altitude target, which can be applied to the field of radar signal processing.

[0035] See also Figure 1 , a method for estimating the elevation angle of a low-altitude target, comprising the following steps:

[0036] Step 1: Obtain radar echo signals, perform coherent accumulation on the array signals, and output the accumulated data.

[0037] Step 2: Separate the accumulated data according to the different terrain environments in the scanning direction, that is, divide the accumulated data into flat terrain data and undulating terrain data; this facilitates the subsequent selection of different algorithms for signal processing.

[0038] In conventional flat terrain, the specular reflection component accounts for the majority, so it can be regarded as a specular model, and the reflecting surface is ordinary soil. In conventional undulating terrain, the diffuse reflection component accounts for the majority, so it cannot be regarded as a specular model. The undulating terrain data is usually distributed in the interval of [-20°, -0.1°] in the pitch dimension.

[0039] Step 3: For flat terrain data, use the RML algorithm for estimation. Perform conventional beamforming and signal detection on the flat terrain data, and perform maximum likelihood estimation on the array composite steering vector w under multipath conditions to obtain the likelihood function T(θ) and obtain the target elevation angle estimate. The specific implementation steps are as follows:

[0040] Step 3a): Perform beamforming on the flat terrain data and perform signal detection on each array element. Perform spectrum analysis (FFT) on the detected target and detect the spectrum peak to obtain the target Doppler frequency estimation value of different array elements.

[0041] Step 3b): Considering that in actual situations, the distance information of the target is often obtained before the elevation angle is estimated. Combined with the information of the antenna height, the geometric relationship between the direct signal and the multipath signal in the echo is obtained, that is, a connection is established between the target elevation angle θ1 and the multipath signal elevation angle θ2, which can reduce the angle search dimension from two dimensions to one dimension. From the mirror reflection signal reception model, it can be seen that:

[0042] θ2=-arcsin(sin(θ1)+2h r / R d )

[0043] Where h r R is the height of the antenna. d is the distance from the antenna to the target.

[0044] Step 3c): Introduce the prior reflection coefficient ρ (radar fixed parameter) to form the composite steering vector of the array under multipath conditions, abbreviated as w:

[0045] w(θ1)=a(θ1)+ρe -jψ a(θ2)=a(θ1)+ρe -jψ a(-arcsin(sin(θ1)+2h r / R d ))

[0046] Where a(θ1) and a(θ2) represent the steering vectors of the direct signal and the multipath signal respectively, ψ represents the phase difference between the direct signal and the multipath signal at the array reference point ψ = 2πΔR / λ, where λ is the radar emission wavelength, and ΔR represents the distance difference between the direct path and the multipath path.

[0047] Step 3d): Find the complex envelope of the direct wave signal after spectral analysis

[0048]

[0049] The number of time domain snapshots is L, S1(m) represents the complex envelope of the direct signal, where f d represents the Doppler frequency of the target, s1 is the initial gain of the target direct wave signal, and Δt is the sampling interval.

[0050] Step 3e): Form a signal array vector

[0051]

[0052] Where n(m) represents the background noise, which is assumed to be zero-mean circular Gaussian white noise and is uncorrelated with the signal.

[0053] Step 3f): Let P w =w[w H w] -1 w H , the likelihood function T(θ) can be obtained by the idea of ​​maximum likelihood estimation:

[0054]

[0055] Step 3g): The estimated value of the target elevation angle θ1 can be obtained from the likelihood function:

[0056]

[0057] From the above formula, we can see that the RML algorithm only needs to perform a one-dimensional search, and because the composite steering vector replaces the conventional steering vector in free space, the maximum likelihood estimation is similar to the maximum likelihood estimation in the case of a single source, so the amount of calculation can be further reduced. In addition, due to the use of the prior information of the reflection coefficient ρ, the angle measurement error of the RML algorithm used has obvious advantages over other algorithms (such as Figure 2 ), the radar angle measurement performance is further improved.

[0058] Step 4: For undulating terrain data, use a multi-frequency beam scanning algorithm. In an undulating ground environment, the specular reflection component in the multipath reflection signal is weak, while the diffuse reflection component plays a major role. At this time, the conventional estimation algorithm has a large error, and a multi-frequency beam scanning method is required for accuracy correction.

[0059] For the target echo data x1 and x2 obtained at two working frequencies, the likelihood functions T1(θ) and T2(θ) are calculated based on their array steering vectors a1(θ) and a2(θ):

[0060]

[0061] Get the target elevation angle estimate

[0062]

[0063] The angle measurement error of the proposed method is significantly reduced compared with the traditional method (e.g. Figure 6 and Fig. 9 ), the radar angle measurement performance is further improved, and the average angle measurement error is less than 0.08°.

[0064] Step 5: Integrate all target elevation angle estimates, combine the elevation angle estimates for flat terrain and undulating terrain, and obtain the elevation angle measurement results within the entire radar scanning range.

[0065] Figure 2-Figure 6 The simulation parameters are as follows: the antenna array is a vertical uniform linear array, the number of array elements is N = 30, the array element spacing is d = 0.125m, the wavelength is λ = 0.23m, and the antenna height is h r =2m, the distance between the target and the radar antenna is R d =100km, reflection coefficient is ρ=-0.9, and the target height is assumed to be 4000m. Each experimental value is obtained from the statistics of 200 independent experimental results.

[0066] Figure 2The RML algorithm used in the present invention has a significant advantage in RMSE compared to other algorithms, which fully proves that the algorithm has good estimation accuracy. Figure 3 It can be clearly seen that when the antenna height is low (such as h r =2m or h r =3m), the theoretical value is basically consistent with the experimental value, and with the increase of antenna height, the overall downward trend of the theoretical value of angle measurement RMSE is obvious. However, when the antenna height is too high, the virtual array produces blurring, so h r =5m, the theoretical value does not match the experimental value. Figure 4 and Figure 5 From the displayed data, it can be seen that when the phase of the reflection coefficient ρ is accurate but the amplitude has an error, the amplitude error has little effect on the angle measurement result; when the amplitude of the reflection coefficient ρ is accurate but the phase has an error, the phase error has a greater impact on the angle measurement result, which proves that the phase of the reflection coefficient ρ has a greater impact on the angle measurement result of the RML algorithm. Figure 6 It is proved that in the case of undulating terrain, the multi-frequency beam scanning algorithm has higher angle measurement accuracy than the single-frequency beam scanning.

[0067] Figure 7-Figure 9 Simulation parameters: The antenna array is a planar array, and the direction parallel to the y-axis is the row direction of the array. The number of array elements in the row direction is 36; the number of array elements in the column direction is 48. The antenna array is tilted 5° and the center height is 1m. The azimuth of the target Slope distance R d =100km, each experimental value is obtained by statistics of 500 independent experimental results. Figure 7 and Figure 8 It can be seen that the RML algorithm has higher measurement accuracy in flat terrain, and the angle measurement error gradually decreases with the increase of signal-to-noise ratio and target elevation angle. Fig. 9 This once again verifies the superiority of the multi-frequency beam scanning algorithm over the single-frequency beam scanning under undulating ground conditions.

[0068] Figure 10-12 The measured parameters are: signal bandwidth 2MHz; data rate 2Msps, each distance unit corresponds to 75m; azimuth windowing for collection, window range 90°-120°, collection weather is clear, temperature 22℃, position wind speed 6; humidity 40%; air pressure 914hpa; a total of 30 rows of antennas in the vertical dimension, the first row of data is the data of the bottom row of antennas, the antenna array is tilted back 15 degrees, and the array channel amplitude and phase errors have been compensated. Scan time: 5s; the horizontal beam width of the array is 2.8 degrees; the center height of the antenna is 5 meters. It can be seen from the self-test data result chart that the RML scanning curve of the measured data is basically consistent with the theoretical RML scanning curve, and when the signal-to-noise ratio is greater than 26dB, the root mean square error of the measured elevation angle is less than 0.2 degrees.

[0069] Based on the above experimental results, it can be seen that the present invention can comprehensively utilize the RML algorithm and the multi-frequency beam scanning algorithm in combination with different ground terrain conditions. In a flat terrain environment, the RML algorithm is used, and the target elevation angle is estimated in combination with the composite steering vector under multipath conditions. This estimation method utilizes a variety of prior information and has the characteristics of low computational complexity and high angular measurement accuracy. In an undulating terrain environment, a multi-frequency beam scanning algorithm is used, which can effectively avoid the angular measurement accuracy interference caused by the strong ground diffuse reflection component, thereby further improving the estimation accuracy.

[0070] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for estimating the elevation angle of a low-altitude target, characterized in that: The steps include: Step 1: Obtain radar echo signals, perform coherent accumulation on array signals, and output the accumulated data; Step 2: According to different terrain environments in the scanning direction, the accumulated data is divided into flat terrain data and undulating terrain data; Step 3: For flat terrain data, use the RML algorithm for estimation; Step 3a): Perform beamforming on the flat terrain data, perform signal detection on each array element, perform spectrum analysis (FFT) on the detected target and detect the spectrum peak to obtain the target Doppler frequency estimation value of different array elements. Step 3b): Establish the connection between the target elevation angle θ1 and the multipath signal elevation angle θ2. The specific formula is: <h2 style=";text-align:left;direction:ltr">θ2 = -arcsin(sin(θ1)+2h<h2 style=";text-align:left;direction:ltr"> r <h2 style=";text-align:left;direction:ltr"> / R<h2 style=";text-align:left;direction:ltr"> d <h2 style=";text-align:left;direction:ltr"> ) Where h r R is the height of the antenna. d is the distance from the antenna to the target; Step 3c): Introduce the prior reflection coefficient ρ to form the composite steering vector of the array under multipath conditions, which is abbreviated as w: w(θ1)=a(θ1)+ρe -jψ a(-arcsin(sin(θ1)+2h r / R d )) Where a(θ1) and a(θ2) represent the steering vectors of the direct signal and the multipath signal, respectively, and ψ represents the phase difference between the direct signal and the multipath signal at the array reference point; Step 3d): Complex envelope of the direct wave signal after spectral analysis The number of time domain snapshots is L, S1(m) represents the complex envelope of the direct signal, and f d Indicates the Doppler frequency of the target; Step 3e): Form a signal array vector Where n(m) represents the background noise; Step 3f): Let P w =w[w H w] -1 w H , likelihood function T(θ): Step 3g): The estimated value of the target elevation angle θ1 obtained by the likelihood function is: Step 4: For undulating terrain data, a multi-frequency beam scanning algorithm is used. The specific steps are as follows: For the target echo data x1 and x2 obtained at two working frequencies, the likelihood functions T1(θ) and T2(θ) are calculated based on their array steering vectors a1(θ) and a2(θ): Get the target elevation angle estimate 2. The method for estimating the elevation angle of a low-altitude target according to claim 1, wherein: The method also includes step 5: integrating all target elevation angle estimation values, combining elevation angle estimation under flat terrain and undulating terrain conditions, and obtaining elevation angle measurement results within the entire radar scanning range.

3. A method for estimating the elevation angle of a low-altitude target according to claim 1 or 2, characterized in that: The phase difference ψ = 2πΔR / λ, where λ is the radar transmission wavelength and ΔR represents the distance difference between the direct path and the multipath path.

4. A method for estimating the elevation angle of a low-altitude target according to claim 1 or 2, characterized in that: f d represents the Doppler frequency of the target, s1 is the initial gain of the target direct wave signal, and Δt is the sampling interval.

5. A method for estimating the elevation angle of a low-altitude target according to claim 1 or 2, characterized in that: The undulating terrain data are distributed in the interval of [-20°,-0.1°] in the pitch dimension.

Images