A method for relocating a moving target inside a building based on an unmanned aerial vehicle dual-channel through-wall radar interferometric phase

Abstract

The application discloses a kind of building interior moving target relocation methods based on unmanned aerial vehicle dual-channel wall-penetrating radar interference phase, comprising: give a new method of moving target azimuth relocation based on dual-channel interference phase.The method first extracts the dual-channel interference phase of moving target in range-compressed domain.Second, establish interference phase polynomial model, and calculate each order coefficient using least square method.Finally, the polynomial coefficient obtained is used to construct the optimization problem about the azimuth position of moving target and solve it.Compared with the traditional method, the application does not need imaging and multi-channel registration operation.In addition, the method proposed in the application can directly obtain the azimuth position of moving target, without estimating radial velocity and calculating offset, avoiding the problem of spatial doppler ambiguity faced by traditional synthetic aperture radar moving target positioning technology in radial velocity estimation.The simulation experiment proves the effectiveness of the proposed method.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing, and more specifically to a method for relocating moving targets using UAV-borne through-wall radar. Background Technology

[0002] In urban warfare and disaster relief scenarios, accurately obtaining the target's location behind obstacles is crucial for tactical planning, gaining operational advantages, and improving rescue efficiency. However, the unknown radial velocity of a moving target can cause image position deviation, making it impossible to directly obtain the true location through imaging. This invention proposes a technique for relocalizing moving targets inside buildings, aiming to accurately estimate the orientation and location of moving targets within obscured spaces.

[0003] Traditional multi-channel moving target azimuth relocation methods include Velocity Synthetic Aperture Radar (VSAR), Along-track Interferometry (ATI), Adaptive Matched Filter (AMF), and Space Projection (SP). These methods all utilize multi-channel moving target phase estimation to determine the radial velocity of the moving target and calculate the azimuth position offset to achieve relocation.

[0004] However, traditional methods for estimating the radial velocity of moving targets require strict registration of multi-channel images. The introduction of a joint pixel signal model, utilizing multi-channel auxiliary pixel information and combining it with traditional radial velocity estimation methods, can achieve accurate estimation of motion parameters where channel registration errors exist. However, such methods rely on a high number of channels and a large degree of spatial freedom, limiting estimation accuracy in dual-channel scenarios.

[0005] Based on the above research, we believe that most existing methods utilize multi-channel moving target phase information to estimate radial velocity and calculate azimuth offset for relocation. This requires the following conditions: the platform velocity is much greater than the moving target velocity, and the shortest slant range is much greater than the moving target's azimuth coordinates. Some methods also require channel registration conditions for the Displaced Phase Center Antenna (DPCA). In through-wall detection scenarios, the azimuth position and slant range histories, as well as the platform velocity and moving target velocity, are all on the same order of magnitude, leading to reduced accuracy in traditional moving target relocation methods. To address these issues, we propose a novel moving target azimuth relocation method. Our method utilizes interferometric phase function extraction and fitting, establishes an optimization problem, and employs pattern search optimization techniques to obtain the moving target's azimuth position, thereby improving the positioning accuracy of the moving target. Summary of the Invention

[0006] This invention provides a moving target relocation method based on dual-channel interferometric phase of an unmanned aerial vehicle (UAV) through-wall radar, comprising:

[0007] Step S1: Acquire the raw echo data of the human target from the UAV-borne through-wall radar and perform pulse compression along the range direction; extract the interferometric phase history of the moving target in the range compression domain;

[0008] Step S2: Derive the functional relationship between the discretized interference phase and slow time, construct a polynomial regression model, and estimate the coefficients of each order using the least squares method;

[0009] Step S3: Construct an optimization function for the orientation and position, and solve it using pattern search.

[0010] Radar transmits linear frequency modulated (LFM) continuous wave signals, i.e.

[0011]

[0012] Where τ represents the fast time variable, f0 is the center frequency, and K r To adjust the frequency, T p Let be the pulse width, and rect(·) be the window function. After scattering by a moving target behind a wall, the echo received by the m-th receiving antenna is:

[0013]

[0014] The received echo signal, after demodulation to baseband, takes the following form:

[0015]

[0016] Where, σ t Let be the scattering coefficient of the moving target, and c be the speed of light.

[0017] After distance pulse compression, we can obtain

[0018]

[0019] Where B = K r T p R represents the bandwidth of the transmitted signal, and λ represents the wavelength. 0m (t a () represents the two-way distance of electromagnetic wave propagation.

[0020] R 0m (t a )=R(t a )+R m (t a )+R wall,0m (5)

[0021] Among them, R m (t a Let be the instantaneous distance from the m-th receiving antenna to the moving target.

[0022]

[0023] R(t a () represents the instantaneous distance from the transmitting antenna to the moving target:

[0024]

[0025] Where x(t) a ),y(t a ) represent the transmitting antenna in slow time t a The horizontal position at time t, and x(t) a ) = V p t a y(t) a ) = 0. Corresponding to slow time t a =Instantaneous slant range from the transmitting antenna to the moving target at time 0. R wall,0m This represents the extra distance electromagnetic waves travel inside the wall.

[0026] The angle between the incident direction of the emitted electromagnetic wave and the plane of the wall is

[0027]

[0028] The angle between the direction of the electromagnetic wave returning from the moving target and the receiving antennas 1 and 2 and the wall is...

[0029]

[0030] but

[0031]

[0032] Where, d w Represents wall thickness, ∈ r This represents the dielectric constant of the wall. Specifically, R corresponds to the free-space scenario. wall,0m =0.

[0033] Distance expressions (6) and (7) in t a Performing a Taylor expansion at time 0 yields:

[0034]

[0035] Then equation (5) can be expanded as follows:

[0036]

[0037] in,

[0038] From formula (13), we can obtain the Doppler frequency of the moving target as follows:

[0039] Among them, f mov The Doppler frequency shift is caused by the radial velocity of the moving target.

[0040]

[0041] v r It is the radial velocity of the moving target:

[0042]

[0043] f p The Doppler frequency is caused by the platform's motion and the azimuth position of the receiving antenna.

[0044]

[0045] f wall The Doppler frequency shift is caused by wall refraction.

[0046]

[0047] K a,mov Frequency adjustment for the azimuth of a moving target:

[0048]

[0049] K a,p Frequency tuning for azimuth caused by radar platform movement:

[0050]

[0051] Specifically, for stationary targets, f mov =0,K a,mov =0, after azimuth compression, it can be accurately focused to the true azimuth position. Compared with the stationary target, the Doppler center of the moving target has shifted. According to (17), the influence of the wall on the shift can be ignored, and only the shift caused by the radial velocity f needs to be compensated. mov According to equation (15), f mov radial velocity v of the moving target r It is related to both the orientation and location x0.

[0052] When the condition R0 >> x0 is satisfied, The azimuth offset is

[0053]

[0054] This allows us to obtain the true location.

[0055] x0=x image -Δx (21) In scenarios involving wall penetration, the condition R0 >> x0 cannot be satisfied, leading to systematic errors in traditional relocation methods.

[0056] This invention performs conjugate multiplication of the results from the two receiving channels in the distance compression domain to obtain...

[0057]

[0058] The interference phase function can be written as

[0059]

[0060] As can be seen from (23), the constant term of the interference phase function includes the azimuth information of the moving target and the additional distance information of the electromagnetic wave propagating in the wall. The component caused by the wall in the interference phase is:

[0061]

[0062] in,

[0063]

[0064] In a typical wall-penetrating scenario, the order of magnitude of B0 is 10. -4 The slant distance difference between the two channels in the constant term of equation (23) The order of magnitude is 10 -2 Therefore, B0 can be ignored here.

[0065] Therefore, the expression for the interference phase is rewritten as follows:

[0066]

[0067] As can be seen from equation (25), after the conjugate multiplication of adjacent channels, the influence of the wall has been eliminated, and the interference phase constant term is only related to d, λ, R0, and x0. Here, R0 can be obtained by indexing the slow time t of the image in the distance-compressed domain. a =0 corresponds to a single snapshot data acquisition.

[0068] The advantage of this invention in performing dual-channel interferometric processing in the range compression domain is that it eliminates the need for registration in the image domain; it can directly estimate the azimuth position of the moving target from the interferometric phase information without estimating the radial velocity and calculating the offset.

[0069] Since the extracted interference phase is a slow-time quadratic function, this invention uses a polynomial regression model for fitting. Equation (25) is the nonlinear interference phase model, and its discrete form can be written as:

[0070]

[0071] Where j is a discrete-time variable, N a n[j] represents the number of sampling points in the azimuth direction, and n[j] represents the noise. Its matrix form is as follows:

[0072]

[0073] in,

[0074]

[0075] θ = [A0, A1, A2] T

[0076] N = [n[1],n[2],…,n[N] a ]] T

[0077] in(·) T This represents the matrix transpose operation. θ represents the coefficients of the polynomial, which can be obtained by joint estimation using the least squares method.

[0078]

[0079] To obtain the bearing and position of a moving target, this invention establishes the following optimization problem:

[0080]

[0081] in,

[0082]

[0083] The above optimization problem can be solved using a pattern search algorithm.

[0084] Beneficial effects:

[0085] (1) This invention proposes a method for relocating moving targets inside buildings based on the interferometric phase of an unmanned aerial vehicle (UAV) dual-channel through-wall radar, which differs from traditional synthetic aperture radar (SAR) moving target localization technology. This technology utilizes electromagnetic waves of a specific frequency band as the transmitted signal, which can penetrate common obstacles such as partitions and walls to locate moving targets hidden behind them. This technology can be used in office buildings, residences, and other building scenarios;

[0086] (2) This invention proposes a method for relocating moving targets inside buildings based on the interferometric phase of a dual-channel through-wall radar on a UAV, which is different from the traditional synthetic aperture radar moving target localization technology. The UAV-borne radar used in this technology can be quickly deployed and approach the target area to collect data, without being limited by terrain or obstacles, and has the advantages of wide-range and multi-angle detection;

[0087] (3) This invention proposes a method for relocating moving targets inside buildings based on the interferometric phase of an unmanned aerial vehicle (UAV) dual-channel through-wall radar, which differs from traditional synthetic aperture radar (SAR) moving target localization techniques. This technique, based on the interferometric phase information of moving targets in the range-compressed domain, can accurately estimate the true azimuth position of the moving target. It eliminates the need for imaging operations, thus avoiding the spatial Doppler ambiguity problem in radial velocity estimation faced by traditional SAR moving target localization techniques. Attached Figure Description

[0088] Figure 1 A schematic diagram of a method for relocating moving targets inside a building based on interferometric phase of a UAV-borne dual-channel through-wall radar is provided by the present invention.

[0089] Figure 2 A schematic diagram illustrating a scenario for a method for relocating moving targets inside a building based on interferometric phase of a UAV-borne dual-channel through-wall radar, provided by this invention.

[0090] Figure 3 This invention provides a schematic diagram of a scenario and radar antenna array structure for a method for relocating moving targets inside buildings based on interferometric phase of a UAV-borne dual-channel through-wall radar.

[0091] Figure 4 This is a schematic diagram illustrating the results of a method for relocating moving targets inside buildings based on interferometric phase of a UAV-borne dual-channel through-wall radar, as provided by the present invention. Detailed Implementation

[0092] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0093] like Figure 1 The method for relocalizing moving targets inside buildings based on interferometric phase detection of UAV-borne dual-channel through-wall radar, as shown, includes:

[0094] Step S1: Acquire the raw echo data of the human target from the UAV-borne through-wall radar and perform pulse compression along the range direction; extract the interferometric phase history of the moving target in the range compression domain;

[0095] Step S2: Derive the functional relationship between the discretized interference phase and slow time, construct a polynomial regression model, and estimate the coefficients of each order using the least squares method;

[0096] Step S3: Construct an optimization function for the orientation and position, and solve it using pattern search.

[0097] Assuming the radar transmits a linear frequency modulated (LFM) continuous wave, that is...

[0098]

[0099] Where τ represents the fast time variable, f0 is the center frequency, and K r To adjust the frequency, T p Let be the pulse width, and rect(·) be the window function. After scattering by a moving target behind a wall, the echo received by the m-th receiving antenna is:

[0100]

[0101] The received echo signal, after demodulation to baseband, takes the following form:

[0102]

[0103] Where, σ t Let be the scattering coefficient of the moving target, and c be the speed of light.

[0104] After distance pulse compression, we can obtain

[0105]

[0106] Where B = K r T p R represents the bandwidth of the transmitted signal, and λ represents the wavelength. 0m (t a () represents the two-way distance of electromagnetic wave propagation.

[0107] R 0m (t a )=R(t a )+R m (t a )+R wall,0m (33)

[0108] Among them, R m (t a Let be the instantaneous distance from the m-th receiving antenna to the moving target.

[0109]

[0110] R(t a () represents the instantaneous distance from the transmitting antenna to the moving target:

[0111]

[0112] Where x(t) a ),y(t a ) represent the transmitting antenna in slow time t a The horizontal position at time t, and x(t) a ) = V p t a y(t) a ) = 0. Corresponding to slow time t a =Instantaneous slant range from the transmitting antenna to the moving target at time 0. R wall,0m This represents the extra distance electromagnetic waves travel inside the wall.

[0113] The angle between the incident direction of the emitted electromagnetic wave and the plane of the wall is

[0114]

[0115] The angle between the direction of the electromagnetic wave returning from the moving target and the receiving antennas 1 and 2 and the wall is...

[0116]

[0117] but

[0118]

[0119] Where, d w Represents wall thickness, ∈ r This represents the dielectric constant of the wall. Specifically, R corresponds to the free-space scenario. wall,0m =0.

[0120] Translate the distance expressions (34) and (35) into t a Performing a Taylor expansion at time 0 yields:

[0121]

[0122] Then equation (33) can be expanded as follows:

[0123]

[0124]

[0125] in,

[0126] From formula (41), we can obtain the Doppler frequency of the moving target as follows:

[0127]

[0128] Among them, f mov The Doppler frequency shift is caused by the radial velocity of the moving target.

[0129]

[0130] v r It is the radial velocity of the moving target:

[0131]

[0132] f p The Doppler frequency is caused by the platform's motion and the azimuth position of the receiving antenna.

[0133]

[0134] f wall The Doppler frequency shift is caused by wall refraction.

[0135]

[0136]

[0137] K a,mov Frequency adjustment for the azimuth of a moving target:

[0138]

[0139] K a,p Frequency tuning for azimuth caused by radar platform movement:

[0140]

[0141] Specifically, for stationary targets, f mov =0,K a,mov =0, after azimuth compression, it can be accurately focused to the true azimuth position. Compared with the stationary target, the Doppler center of the moving target has shifted. According to (45), the influence of the wall on the shift can be ignored, and only the shift caused by the radial velocity f needs to be compensated. mov According to equation (43), f mov radial velocity v of the moving target r It is related to both the orientation and location x0.

[0142] When the condition R0 >> x0 is satisfied, The azimuth offset is

[0143]

[0144] This allows us to obtain the true location.

[0145] x0=x image -Δx (49) In scenarios involving wall penetration, the condition R0 >> x0 cannot be satisfied, leading to systematic errors in traditional relocation methods.

[0146] This invention performs conjugate multiplication of the results from the two receiving channels in the distance compression domain to obtain...

[0147]

[0148] The interference phase function can be written as

[0149]

[0150]

[0151] As can be seen from (23), the constant term of the interference phase function includes the azimuth information of the moving target and the additional distance information of the electromagnetic wave propagating in the wall. The component caused by the wall in the interference phase is:

[0152]

[0153] in,

[0154]

[0155] In a typical wall-penetrating scenario, the order of magnitude of B0 is 10. -4 The slant distance difference between the two channels in the constant term of equation (51) The order of magnitude is 10 -2 Therefore, B0 can be ignored here.

[0156] Therefore, the expression for the interference phase is rewritten as follows:

[0157]

[0158] As can be seen from equation (53), after the conjugate multiplication of adjacent channels, the influence of the wall has been eliminated, and the interference phase constant term is only related to d, λ, R0, and x0. Here, R0 can be obtained by indexing the slow time t of the image in the distance-compressed domain. a =0 corresponds to a single snapshot data acquisition.

[0159] The advantage of this invention, which performs dual-channel interferometry in the range compression domain, is that it eliminates the need for registration in the image domain; it allows direct estimation of the moving target's azimuth from the interferometric phase information, without requiring radial velocity estimation and offset calculation.

[0160] Since the extracted interference phase is a slow-time quadratic function, this invention uses a polynomial regression model for fitting. Equation (53) is the nonlinear interference phase model, and its discrete form can be written as:

[0161]

[0162] Where j is a discrete-time variable, N a n[j] represents the number of sampling points in the azimuth direction, and n[j] represents the noise. Its matrix form is as follows:

[0163]

[0164] in,

[0165]

[0166] θ = [A0, A1, A2] T

[0167] N = [n[1],n[2],…,n[N] a ]] T

[0168] in(·) T This represents the matrix transpose operation. θ represents the coefficients of the polynomial, which can be obtained by joint estimation using the least squares method.

[0169]

[0170] To obtain the bearing and position of a moving target, this invention establishes the following optimization problem:

[0171]

[0172] in,

[0173]

[0174] The above optimization problem can be solved using a pattern search algorithm to obtain the final location result.

[0175] This invention proposes a method for relocating moving targets inside buildings based on interferometric phase information from a UAV-borne dual-channel through-wall radar, which differs from traditional synthetic aperture radar (SAR) moving target localization techniques. This technique utilizes electromagnetic waves in a specific frequency band as the transmitted signal, which can penetrate common obstacles such as partitions and walls to locate moving targets hidden behind them. This technology can be applied to office buildings, residences, and other building scenarios. Furthermore, the UAV-borne radar used in this technique has strong flexibility and maneuverability, enabling rapid deployment and proximity to the target area to collect data, unrestricted by terrain or obstacles, and possessing the advantages of wide-range, multi-angle detection. Based on moving target interferometric phase information in the range-compressed domain, this technique can accurately estimate the true azimuth and position of the moving target. No imaging operation is required, avoiding the spatial Doppler ambiguity problem in radial velocity estimation faced by traditional SAR moving target localization techniques. In summary, this invention has significant application value in the field of moving target relocation, providing an effective solution for locating moving targets in obscured spaces.

[0176] like Figure 3 As shown, Figure 3 (a) is a schematic diagram of a simulation scene of a moving human target passing through a wall, with a moving target located behind a brick wall; Figure 3 (b) is a diagram of an unmanned aerial vehicle (UAV) through-wall radar array, including one transmitting antenna and two receiving antennas.

[0177] like Figure 4 As shown, this is in Figure 3 Simulation results in the scenario. Figures (a), (b), (c), (d), and (e) show the relocation results for: moving target with different azimuth velocities; moving target with different range velocities; moving target with different azimuth positions; moving target with different range positions; and relocation results under different signal-to-noise ratios. Compared to traditional methods, the results of this invention demonstrate good positioning accuracy under different signal-to-noise ratios, moving target velocities, and positions, achieving precise azimuth relocation of moving targets behind walls.

[0178] The specific embodiments described above only illustrate the design principles of the present invention. The shapes and names of the components in this description may differ and are not limited. Therefore, those skilled in the art can modify or make equivalent substitutions to the technical solutions described in the foregoing embodiments; and these modifications and substitutions do not depart from the inventive spirit and technical solutions of the present invention, and should all fall within the protection scope of the present invention.

Claims

1. A method for relocating moving targets using an unmanned aerial vehicle (UAV) through-wall radar based on dual-channel interferometric phase detection, characterized in that, include: Step S1: Acquire the raw echo data of the human target from the UAV-borne through-wall radar and perform pulse compression along the range direction; extract the interferometric phase history of the moving target in the range compression domain; Step S2: Derive the functional relationship between the discretized interference phase and slow time, construct a polynomial regression model, and estimate the coefficients of each order using the least squares method; Step S3: Construct an optimization function for the orientation and position, and solve it using pattern search; The interference phase function is （1）； in, Represents wavelength, Corresponding slow time The instantaneous slant range between the transmitting antenna and the moving target. Radial velocity representing a moving target: In equation (1), the constant term of the interference phase function includes the azimuth and position information of the moving target, as well as the additional distance information of the electromagnetic wave propagating in the wall; the component of the interference phase caused by the wall is: （2）； in, ； in, Represents wall thickness; In the scene where you pass through a wall, The order of magnitude is 10⁻⁴, while the constant term in equation (2) contains the difference in slant distance between the two channels. The order of magnitude is 10⁻², therefore here Negligible; The expression for the interference phase is rewritten as follows （3）； in, This represents the two-way distance of an electromagnetic wave propagating from the transmitting antenna to the moving target and then to the m-th receiving antenna. In equation (3), after the conjugate multiplication of adjacent channels, the influence of the wall has been eliminated, and the interference phase constant term is only related to... , and related; Slow time for image indexing by distance from the compressed domain Data acquisition for a single snapshot; Since the extracted interference phase is a slow-time quadratic function, this invention uses a polynomial regression model for fitting; equation (3) is the nonlinear interference phase model, and its discrete form is written as: （4）； in, j For discrete-time variables, Represents the number of sampling points in the azimuth direction. Represents noise; its matrix form is as follows ； in, ； ； ； ； in Represents the matrix transpose operation; The coefficients of each order of the polynomial are obtained by joint estimation using the least squares method: 。 2. The method as described in claim 1, characterized in that, The radar transmits a linear frequency modulated (LFM) continuous wave signal, i.e. （5）； in, Represents fast-time variables. For the center frequency, To adjust the frequency, The pulse width. The window function; scattering from a moving point target behind a wall, the th The echo received by each receiving antenna is: （6）； Where c represents the speed of light. The scattering coefficient is the scattering coefficient of the moving target.

3. The method as described in claim 1, characterized in that, The received echo signal, after demodulation to baseband, takes the following form: （7）； in, Let be the scattering coefficient of the moving target. It is the speed of light.

4. The method as described in claim 1, characterized in that, After distance pulse compression, the following is obtained （8）； in, For the transmission signal bandwidth, Represents wavelength; Two-way distance representing the propagation of electromagnetic waves （9）； in, For the first Instantaneous distance from each receiving antenna to the moving target: （10）； The instantaneous distance from the transmitting antenna to the moving target: （11）； in, The transmitting antenna in slow time The position in the azimuth direction and the position in the distance direction at any given time, and , Corresponding slow time The instantaneous slant range between the transmitting antenna and the moving target at all times; This represents the extra distance electromagnetic waves travel inside the wall.

5. The method as described in claim 1, characterized in that, The angle between the incident direction of the emitted electromagnetic wave and the plane of the wall is （12）； The angle between the direction of the electromagnetic wave returning from the moving target and the receiving antennas 1 and 2 and the wall is... （13）； but （14）； in, Represents wall thickness. Represents the dielectric constant of the wall; specifically, the free space scenario corresponds to... .

6. The method as described in claim 1, characterized in that, Distance expressions (10) and (11) are in Performing a Taylor expansion at each step, we obtain: （15）； （16）； Equation (9) expanded as （17）； in, .

7. The method as described in claim 1, characterized in that, From formula (17), the Doppler frequency of the moving target is obtained as follows: （18）； in, The Doppler frequency shift is caused by the radial velocity of the moving target. （19）； It is the radial velocity of the moving target: ； The Doppler frequency is caused by the platform's motion and the azimuth position of the receiving antenna. （20）； The Doppler frequency shift is caused by wall refraction. （21）； Frequency adjustment for the azimuth of a moving target: （22）； Frequency tuning for azimuth caused by radar platform movement: （23）； In particular, for stationary targets, , After azimuth compression, it can be accurately focused to the true azimuth position; compared with the stationary target, the Doppler center of the moving target has shifted; according to (21), the influence of the wall on the Doppler frequency shift of the moving target is 0, and only the shift caused by the radial velocity needs to be compensated. According to equation (19), radial velocity of the moving target and location Both are related.

8. The method as described in claim 1, characterized in that, When satisfied Under what conditions, ; The azimuth offset is （24）； This allows us to obtain the true location. （25）； In scenarios involving passing through walls, it cannot satisfy... These conditions lead to systematic errors in traditional relocation methods.

9. The method as described in claim 1, characterized in that, Multiplying the results from the two receiving channels conjugate in the distance compression domain yields... （26）。 10. The method as described in claim 1, characterized in that, To obtain the bearing and position of a moving target, the following optimization problem is established: （27）； in, （28）。 11. The method as described in claim 1, characterized in that, Equation (27) is solved using a pattern search algorithm, and the final location of the moving target is estimated.

Images