A moving target length estimation method based on ultra-wideband signal

Abstract

The present invention belongs to the field of radar identification technology, and specifically relates to a moving target length estimation method based on ultra-wideband signals. The present invention proposes a length estimation method for targets that are simultaneously performing micro-motion and high-speed motion based on ultra-wideband electromagnetic scattering characteristics, making full use of known information to achieve a more accurate estimation of the target structure information. Compared with the existing models that only consider the high-speed motion of the target or only consider the micro-motion of the target, the target motion model in this method simultaneously considers the state of the target performing micro-motion and high-speed motion, achieving a target length estimation that is closer to reality, and at the same time can verify that the estimation method has better accuracy than ignoring the high-speed motion of the target in actual situations.

Description

Technical Field

[0001] The invention belongs to the technical field of radar identification, and in particular relates to a moving target length estimation method based on ultra-wideband signals. Background Art

[0002] From the principle of radar imaging, we can know that the bandwidth of the radar (transmitted signal) has a significant impact on the range resolution of the radar. If the range resolution is to be improved, the signal bandwidth needs to be greatly increased. Ultra-wideband refers to the relative bandwidth of the signal. When the ratio of the bandwidth of the signal to the center frequency is greater than 25%, it is called an ultra-wideband (UWB) signal, between 1% and 25% is wideband (WB), and the ratio of bandwidth to center frequency is less than 1% is called narrowband (NB). Ultra-wideband radar is currently a new system radar. Many relevant target information exists in the electromagnetic echo of the target. Compared with traditional system radar, many signal analysis methods can be used to obtain more accurate target features and extract more reliable target features than traditional system radar.

[0003] In the military field, ballistic missiles are undoubtedly one of the most threatening weapons in war due to their high precision, great power and range of tens of thousands of kilometers. They have an impact that cannot be ignored. At present, with the development of science and technology, including the advancement of war forms, ballistic missile manufacturing has spread all over the world. In order to maintain national security, research on ballistic missile defense needs to be more extensive and in-depth. Missile defense systems are generally composed of early warning detection systems, interception weapon systems, and combat management and communication systems. Missiles are usually divided into three stages during flight: boost phase, mid-stage, and reentry phase. Among them, the longest stage of missile flight in the target orbit is the mid-stage. In the early warning system, the most common is to detect targets in the mid-stage process, distinguish them from decoys, and conduct threat assessment of warheads through identification to make subsequent interception actions. The identification of warheads has always been a core problem. At present, the distinction between true and false warheads has put forward higher requirements for sensors and signal analysis capabilities in the early warning system.

[0004] The warhead is flying at a high speed in the actual flight process. In the mid-stage, due to separation from the mother cabin, the warhead will be affected by the lateral torque, resulting in the precession phenomenon unique to warhead targets. When extracting the characteristics of the target, we cannot ignore the influence of its macroscopic and microscopic motion. Therefore, it is necessary to analyze the warhead in the motion model and study the target echo characteristics obtained based on ultra-wideband radar to obtain more detailed characteristics of the target as much as possible for the threat assessment stage of missile defense research, so as to make the most reasonable defense decision.

[0005] The high range resolution of broadband radar gives it many characteristics and advantages in target detection. Through the rich electromagnetic echo information it carries, more robust and reliable target features can be extracted, and the high-resolution one-dimensional range profile (HRRP) can obtain the target's structural information. The target broadband one-dimensional range profile is equivalent to the projection of the three-dimensional scattering point corresponding to the target on the radar ray, which reveals the distribution of the target's scattering intensity along the line of sight, reflects the fine structural characteristics of the target, and is a good target recognition feature. Although existing broadband radar systems can provide high-range resolution, the fine structures of missiles, aircraft, and satellites are usually smaller than the range resolution of broadband radar systems. Therefore, it is difficult for existing broadband radar systems to extract the fine structures of such targets, and the establishment of a physical UWB radar may lead to increased complexity in system design and a significant increase in hardware costs. An effective option is to synthesize UWB data from multiple frequency sub-bands without significantly improving the hardware of existing radar systems. The synthesis of multiple frequency sub-bands can be achieved based on the geometric theory of diffraction (GTD) model. Due to its high computational accuracy and relatively simple calculation process, the GTD model has been studied and applied in electromagnetic scattering problems for decades. Summary of the invention

[0006] The present invention proposes a method for estimating the length of a target that is simultaneously moving micro-motion and high-speed motion based on ultra-wideband electromagnetic scattering characteristics, which makes full use of known information to achieve a more accurate estimation of the target structure information. Compared with the existing model that only considers the high-speed motion of the target or only considers the micro-motion of the target, the target motion model in this method simultaneously considers the state of the target moving micro-motion and high-speed motion, achieving a target length estimation that is closer to reality, and at the same time, it can be verified that the estimation method has better accuracy than ignoring the high-speed motion of the target in actual situations.

[0007] The solution of the present invention is: first, establish a motion model of the target, and analyze the influence of the change of the target's attitude angle on the relative distance between the scattering centers on the target. Secondly, according to the geometric theory of the target diffraction model, synthesize multiple frequency sub-bands and parameter characteristics, wherein the parameter characteristics include relative range, amplitude, frequency response parameters and the number of target scattering centers. Then, the echo signal distortion caused by the high-speed movement of the target is compensated for by speed, and then the undistorted HRRP is obtained by inverse Fourier transform. Finally, when the target micro-motion cycle and micro-motion angle change are known, the target attitude angle change within the cycle is searched, and the target length information is analyzed and calculated by combining the angle between the target radial and the radar line of sight with the change of the relative distance of the scattering centers in the target HRRP.

[0008] The technical solution of the present invention is:

[0009] Step 1: Establish a motion model for a moving target with known precession angle and micromotion period, obtain the change in relative distance of the scattering center on the target on the radar line of sight within a micromotion period, and conduct preliminary analysis.

[0010] Step 2: According to the transmission signals of radars in different bands and the motion information of the observed target, a target echo signal model of the broadband radar based on geometric diffraction is established to obtain the sub-band echo signal (sampling one motion cycle according to the model motion).

[0011] Step 3: Get the estimated speed and compensate the sub-band echo signal. Use the frequency band extrapolation technology to expand the frequency range of the sub-band echo signal to the full band to obtain the extended echo signal.

[0012] Step 4: Perform inverse Fourier transform on the echo signal obtained in step 3 to obtain HRRP.

[0013] Step 5: Obtain the target scattering center relative distance change curve from the HRRP obtained in step 4

[0014] Step 6: Traverse the initial attitude angles that the target can have within a cycle, and find the real attitude angle of the target in combination with the distance change curve obtained in step 5, and obtain the estimated length of the target.

[0015] The invention has the following beneficial effects: the invention is a radar target recognition method based on ultra-wideband signals, the target takes into account both its micro-motion characteristics and high-speed motion, and when the initial attitude angle and high-speed motion speed of the target are unknown, the echo signal is analyzed, and the structural change characteristic information of the target scattering center is obtained by using the sub-band HRRP data information, that is, the target length change curve in the HRRP, and then the angle that the target can be used as the initial attitude is traversed to find the change curve of the radial and radar line of sight angle cosine value of the target within a period, so as to obtain the target length estimation value according to the relationship that the ratio of the two should be the real length of the target. Therefore, the method can estimate the length information of the target for a warhead-like motion model that is closer to the actual one. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the top-level structure of the method of the present invention;

[0017] Figure 2 Schematic diagram of target motion model;

[0018] Figure 3 The trajectory diagram of the two scattering centers when the target precesses within one cycle

[0019] Figure 4 The relative distance of the target scattering center obtained by HRRP on the radar line of sight, that is, the HRRP test length change curve of the target at different attitude angles within one cycle

[0020] Figure 5 These are the cosine variation curves of the target radial direction and the radar line of sight when only considering the target micro-motion and when considering both the target's high-speed motion and micro-motion. DETAILED DESCRIPTION

[0021] The present invention will be further described below in conjunction with the accompanying drawings.

[0022] like Figure 1 As shown, it is the process of the present invention, which specifically includes:

[0023] Step 1: Establish a motion model for a moving target with known precession angle and micromotion period, and analyze the change in the relative distance of the scattering center on the target on the radar line of sight within a micromotion period.

[0024] Step 1-1: First, establish the target micro-motion model, such as Figure 2 As shown in the figure, the radar coordinate system is (U, V, W), and the origin, that is, the radar position, is Q. The reference coordinate system parallel to the radar coordinate system (U, V, W) is (X, Y, Z), which only moves with the target. The origin is the center S of the cone bottom circle. The target coordinate system that describes the information of each point on the target is (x, y, z), with the origin as the center of mass O, the cone with the bottom surface S as the center, SN as the precession axis, and the angular velocity ω c Perform cone rotation, with Oz as the spin axis and the angular velocity ω s Start micro-movement. Let the coordinates of point S in the target coordinate system be (0,0,z0), then the initial position of any scattering point P (x,y,z) on the cone in the reference coordinate system can be expressed as:

[0025]

[0026] Where R is the transformation matrix from the target coordinate system to the reference coordinate system, which is composed of the initial Euler angle Determine the angles between the x-axis and the X-axis, the angle between the y-axis and the Y-axis, and the angle between the z-axis and the Z-axis.

[0027]

[0028] At time t, the radial distance from the scattering point P (x, y, z) to the radar is:

[0029]

[0030] Where R ω is the micro-motion matrix, is the radar line of sight unit vector.

[0031] The micro-motion of warhead-type targets usually includes the spin around its own symmetry axis and the cone rotation around a fixed axis SN in one direction. However, if the target has rotational symmetry, such as a cone target like a warhead, the spin does not modulate the target echo. In this case, the spin angular velocity can be regarded as 0. The cone motion is the change in the position of the scattering center as follows:

[0032] At time t, the transformation matrix caused by the target cone rotation can be expressed as:

[0033]

[0034] in, is the angular velocity change vector, is the antisymmetric matrix of a unit direction vector, that is

[0035]

[0036] According to the Rodrigues rotation equation around a vector:

[0037]

[0038] Where I is the third-order unit matrix, Ω is the angular velocity of rotation, is the cone rotation unit direction vector, and the antisymmetric matrix obtained by the target cone rotation transformation matrix is:

[0039]

[0040] (α N ,β N ) represent the azimuth and elevation angles of the cone rotation axis SN in the reference coordinate system, respectively.

[0041] At time t, the position of the scattering point P can be expressed in the reference coordinate system as:

[0042]

[0043] Step 1-2: High-speed motion is represented by the target moving in a certain direction at a speed (v x ,v y ,v z ), the position of the target bottom center S at time t in the radar coordinate system can be expressed as (x0)v x t,y0)v y t,z0)v z t), where (x0, y0, z0) represents the initial coordinate position of P in the radar coordinate system. This movement will bring about a change in the radar line of sight angle;

[0044] Step 1-3: The relative position of the scattering center P of a point on the radar line of sight considering both micro-motion and high-speed motion can be expressed as in Represents the radar line of sight unit vector at time t.

[0045] Step 2: According to the transmission signals of radars in different bands and the motion information of the observed target, the target echo signal model of broadband radar based on geometric diffraction is established to obtain the sub-band echo signal:

[0046]

[0047] Where y is the radar scattering echo, r m represents the relative distance between the mth scattering center and the reference point, f is the echo signal frequency, f=f0+n·Δf, f0 is the signal starting frequency, Δf represents the frequency interval, and n is the number of sampling points. M represents the order of the model, that is, the number of scattering centers, A m Represents the amplitude, which is the intensity coefficient of the mth scattering center. m represents the frequency dependence factor, that is, the geometric type parameter of the mth scattering center, c is the propagation speed of electromagnetic waves in space, V represents additive complex white Gaussian noise.

[0048] Step 3: In the target speed range v min ~v max , to set the step value v step Calculate the entropy of the target one-dimensional range image after speed compensation, and search for the minimum entropy value in the obtained target one-dimensional range image entropy. The speed corresponding to the minimum entropy value is the estimated speed. where v min , v max are the upper and lower limits of the target speed respectively;

[0049] The one-dimensional range image entropy of the target is:

[0050]

[0051] in is the amplitude distribution of the sampling points, Y i = {Y i (h′)|h′=1,2,...,n′} is y i (f i ) The one-dimensional distance image obtained by inverse fast Fourier transform, y i (f i ) is the echo signal after compensation, f i =f′0+n i Δf i , f′0 represents the starting frequency of the full frequency band, Δfi is the frequency sampling interval of the i-th radar, n1=0,...,N1-1, n2=N′-N2,...,N′, N1, N2 represent the number of frequency steps, N′ is the number of frequency sampling points in the full frequency band, i=1,2, h=1,2,...,L, l=1,2,...,L, L is the length of the speed value interval, and the compensation phase is:

[0052]

[0053] in, is the estimated speed, K i is the frequency modulation slope, f ci is the frequency interval, R m is the relative position of the scattering center.

[0054] Step 4: Perform inverse Fourier transform on the echo signal obtained in step 3 to obtain HRRP.

[0055] Step 5: Obtain the target scattering center relative distance change curve from the HRRP obtained in step 4, and extract the target scattering center length in the HRRP using the threshold judgment method.

[0056] Assuming that the threshold used to screen the scattering center is η, the decision criterion for distinguishing the target scattering center from the noise at the sampling point is as follows:

[0057] |X(k)|>η,k=0,1,...,N″-1

[0058] Where X(k) represents the amplitude of each sampling point of HRRP, there are N″ sampling points in total, let vector p = {k||X(k)|≥η, k = 0, 1, ..., N″-1}, N″ is the number of sampling points, Δf represents the frequency interval of radar signal, c is the speed of light, then the estimated value of target length can be obtained as follows

[0059]

[0060] Step 6: Traverse the initial attitude angles that the target can have within a cycle, add the estimated speed influence in step 3, and combine the distance change curve obtained in step 5 to find the target's true attitude angle, and obtain the estimated length of the target. Since the length of the target in HRRP is the projection length of the target radial on the radar line of sight, the ratio of the test length to the cosine value of the angle between the target radial and the radar line of sight, L', should be the target's true length L. In the actual process, the target's initial attitude angle is unknown. By traversing each possible attitude angle, substituting it into the motion model, and then calculating the target's true length, the calculation results at each moment should be the same, so we can consider the minimum mean square error of the calculated length within a cycle as the target's initial attitude angle, and then average the estimated length L' obtained within a cycle, which is considered to be the final target estimated length.

[0061] In order to verify the correctness of the solution of this method, the present invention has carried out simulation experiments. The target radar signal is LFM starting at 3GHZ, with a bandwidth of 1GHZ and a carrier frequency of 5GHZ. The GTD model is used for simulation. The target mass center is located at (4000, 9000, 7000) in the reference coordinate system. At the start time, in the target coordinate system, the target vertex position is (0, 0, 1.2), the two scattering centers are vertex and P (-0.1, 0.1, -0.6), the target bottom center S coordinate is (0, 0, -0.6), the target initial Euler angle is (20°, 30°, 45°), the precession axis is in the reference coordinate system. The azimuth angle is (60°, 45°), the angular velocity is 4πrad / s, and the period is 0.5 seconds.

[0062] Figure 2 The trajectory diagram of the two scattering centers within one period of the target also shows that the target precession trajectory conforms to the theoretical setting. Figure 3 That is, the relative distance change of the target scattering center obtained by HRRP within the period, that is, the test length. Figure 4 In order to consider whether the target is moving at high speed or not, the cosine value change curve of the target radial and the radar line of sight angle is obtained. It can be seen that there is a certain difference between the two. The impact of high-speed movement cannot be ignored in the actual process.

[0063] Experimental results: Table 1 shows the mean and variance of the estimated length considering whether the target is moving at high speed:

[0064] Table 1 The impact of target motion model establishment on length estimation

[0065] Target motion model Consider only micro-movement Simultaneous high-speed movement and micro-motion Mean 1.7 1.74 variance 0.15 0.03

[0066] By comparison, it can be found that in actual situations, after considering both high-speed motion and micro-motion of the target, the initial attitude angle of the target is estimated, and the variance of the test length sequence obtained is smaller, and the mean is closer to the true value of 1.8m. Therefore, the model and estimation method established in this paper are more accurate than the results obtained by considering only one type of motion, and the estimated value of the target length is not much different from the true value, which meets the experimental objectives.

Claims

1. A moving target length estimation method based on ultra-wideband signal, characterized in that: The following steps are involved: Step 1: Establish a motion model for a moving target with known precession angle and micromotion period, and obtain the change in relative distance of the scattering center on the target in the radar line of sight within a micromotion period; the specific method is: Step 11: Establish the target micro-motion model: define the radar coordinate system as (U, V, W), and the origin, i.e. the position of the radar, as Q; the reference coordinate system parallel to the radar coordinate system (U, V, W) is (X, Y, Z), which only moves with the target, and the origin is the center S of the cone bottom circle. The target coordinate system that describes the information of each point on the target is (x, y, z), with the origin as the center of mass O, the cone centered on the bottom surface S, SN as the precession axis, and the angular velocity ω c Perform cone rotation, with Oz as the spin axis and the angular velocity ω s Start micro-movement; let the coordinates of point S in the target coordinate system be (0,0,z0), then the initial position of any scattering point P (x,y,z) on the cone in the reference coordinate system is expressed as: Where R is the transformation matrix from the target coordinate system to the reference coordinate system, which is composed of the initial Euler angle Sure, They are the angles between the x-axis and the X-axis, the angles between the y-axis and the Y-axis, and the angles between the z-axis and the Z-axis: At time t, the radial distance from the scattering point P (x, y, z) to the radar is: Where R ω is the micro-motion matrix, is the radar line of sight unit vector; The spin angular velocity of the warhead target is considered to be 0, and the conical motion is the change of the scattering center position as follows: At time t, the transformation matrix caused by the target cone rotation is expressed as: in, is the angular velocity change vector, is the antisymmetric matrix of a unit direction vector, that is According to the Rodrigues rotation equation around a vector: Where I is the third-order unit matrix, Ω is the angular velocity of rotation, is the cone rotation unit direction vector, and the antisymmetric matrix obtained by the target cone rotation transformation matrix is: (α N ,β N ) represent the azimuth and elevation angles of the cone rotation axis SN in the reference coordinate system, respectively; At time t, the position of the scattering point P is expressed in the reference coordinate system as: Step 12: Express high-speed motion as the target moving in a certain direction at a speed (v x ,v y ,v z ), the position of the target bottom center S at time t in the radar coordinate system is expressed as (x0)v x t,y0)v y t,z0)v z t), where (x0, y0, z0) represents the initial coordinate position of P in the radar coordinate system. This movement will bring about a change in the radar line of sight angle; Step 13: The relative position of the scattering center P of a point on the target considering both micro-motion and high-speed motion on the radar line of sight is expressed as in represents the radar line of sight unit vector at time t; Step 2: Based on the motion model established in step 1, according to the transmission signals of radars in different bands and the motion information of the observed target, establish the target geometric diffraction echo signal model of the broadband radar to obtain the sub-band echo signal: Where y is the radar scattering echo, r m represents the relative distance between the mth scattering center and the reference point, f is the echo signal frequency, f = f0)n·Δf, f0 is the signal starting frequency, Δf represents the frequency interval, n is the number of sampling points, M represents the order of the model, that is, the number of scattering centers, A m represents the amplitude, which is the intensity coefficient of the mth scattering center, α m represents the frequency dependence factor, that is, the geometric type parameter of the mth scattering center, c is the propagation speed of electromagnetic waves in space, V represents additive complex Gaussian white noise; Step 3: In the target speed range v min ~v max , to set the step value v step Calculate the entropy of the target one-dimensional range image after speed compensation, and search for the minimum entropy value in the obtained target one-dimensional range image entropy. The speed corresponding to the minimum entropy value is the estimated speed. where v min , v max They are the upper and lower limits of the target speed respectively; The one-dimensional range image entropy of the target is: in is the amplitude distribution of the sampling points, Y i = {Y i (h′)|h′=1,2,...,n′} is y i (f i ) The one-dimensional distance image obtained by inverse fast Fourier transform, y i (f i ) is the echo signal after compensation, f i =f0′+n i Δf i , f0′ represents the starting frequency of the full frequency band, Δf i is the frequency sampling interval of the i-th radar, n1=0,...,N1-1, n2=N′-N2,...,N′, N1 and N2 represent the number of frequency steps, N′ is the number of frequency sampling points in the full frequency band, i=1,2, h=1,2,...,L, l=1,2,...,L, L is the length of the speed value interval, and the compensation phase is: in, is the estimated speed, K i is the frequency modulation slope, f ci is the frequency interval, R m is the relative position of the scattering center; Step 4: Perform inverse Fourier transform on the echo signal obtained in step 3 to obtain HRRP; Step 5: Obtain the target scattering center relative distance change curve from the HRRP obtained in step 4, and extract the target scattering center length in the HRRP using the threshold judgment method: Assuming that the threshold used to screen the scattering center is η, the decision criterion for distinguishing the target scattering center from the noise at the sampling point is as follows: |X(k)|>η,k=0,1,...,N″-1 Where X(k) represents the amplitude of each HRRP sampling point, there are N″ sampling points in total, let vector p = {k||X(k)|≥η, k = 0, 1, ..., N″-1}, Δf represents the frequency interval, c is the speed of light, and the estimated value of the target length is obtained as follows: Step 6: Traverse the initial attitude angle of the target within a cycle, add the estimated speed influence in step 3, and combine the distance change curve obtained in step 5 to find the real attitude angle of the target and obtain the estimated length of the target.

Images