A moving target detection method based on spatio-temporal-frequency transformation space slices

Abstract

The present invention belongs to the field of underwater target detection, and discloses a moving target detection method based on spatio-temporal-frequency transformation space slicing. The target radiated acoustic signals received by an M-element horizontal line array in a low signal-to-noise ratio underwater acoustic environment are segmented; the received signals on each array element for each period of time are respectively subjected to N-point DFT; frequency-domain beamforming is performed on the array signals after each period of DFT, and after compensating for the phase difference between the arrays brought by the azimuth angle of each element, they are superimposed; coordinate transformation is performed on the obtained three-dimensional matrix space of frequency-azimuth-time; slices are taken in the obtained three-dimensional space of frequency-azimuth-time after coordinate transformation; and segmented Radon transformation is performed on the obtained space slices to detect the target. The present invention is used to solve the passive detection of uniformly moving targets in a low signal-to-noise ratio environment, especially for targets with relatively rapid changes in azimuth angle.

Description

Technical Field

[0001] The present invention belongs to the field of underwater target detection, and particularly relates to a moving target detection method, an electronic device and a readable storage medium based on space-time-frequency transformation space slicing. Background Art

[0002] In a low signal-to-noise ratio underwater acoustic environment, traditional signal detection methods severely restrict the detection performance of underwater sonar arrays. In recent years, the method of using multi-frame data accumulation to improve the signal-to-noise ratio of target signals has become a research hotspot. A large number of design analyses and computer simulations have demonstrated the advantages of the track-before-detect algorithm. However, currently only a limited number of finished systems are applied to actual work. One of the main reasons is that as the integration time (number of frames) increases, the change in signal parameters caused by target movement cannot be ignored. Therefore, in the case where the target movement parameters are unknown, it is impossible to accurately integrate the signal for a long time.

[0003] In the research work of improving the signal-to-noise ratio based on long-time integration in the underwater acoustic field, the focus is on the discussion of target parameters in the frequency domain or spatial domain. Most of them only consider the frequency change of moving targets alone or only conduct target movement analysis for pure azimuth angles. It was not until recent years that the processing and analysis method of space-time-frequency joint was started to be discussed. However, the vast majority of solutions only consider the analysis of target movement parameters in a high signal-to-noise ratio environment. Because for moving targets in a low signal-to-noise ratio environment, the search problem of multi-dimensional movement parameters needs to be solved. So far, there is still a lack of a simple, fast and effective method for multi-frame accumulation of low signal-to-noise ratio moving targets. Summary of the Invention

[0004] The present invention provides a moving target detection method based on space-time-frequency transformation space slicing to solve the passive detection of uniformly moving targets in a low signal-to-noise ratio environment, especially for targets with relatively fast azimuth angle changes.

[0005] The present invention provides an electronic device for running the method steps in a computer-readable storage medium.

[0006] The present invention provides a computer-readable storage medium for storing the method steps of the moving target detection method based on space-time-frequency transformation space slicing.

[0007] The present invention is realized through the following technical solutions:

[0008] A moving target detection method based on space-time-frequency transformation space slicing, the moving target detection method includes the following steps:

[0009] Step 1: Segment the target radiated acoustic signal s(t) received by an M-element horizontal line array in a low signal-to-noise ratio underwater acoustic environment;

[0010] Step 2: For each period τ in Step 1 p Perform N-point DFT on the received signals at each array element, where N = T0·f s ;

[0011] Step 3: Perform frequency-domain beamforming on the array signals after DFT in Step 2, and superimpose them after compensating for the phase difference between array elements caused by the azimuth angle α(τ p );

[0012] Step 4: Perform coordinate transformation on the frequency-azimuth-time (f-α-t) three-dimensional matrix space obtained in Step 3;

[0013] Step 5: Take slices in the frequency-azimuth-time (f-cosθ-t) three-dimensional space obtained by the coordinate transformation in Step 4;

[0014] Step 6: Perform segmented Radon transform on the spatial slices obtained in Step 5 to detect targets.

[0015] A moving target detection method based on spatial slices of spatio-temporal frequency transformation. Specifically, Step 1 is as follows

[0016] s m (τ p ,n) = s((τ p - 1)T b + n) (1)

[0017] τ p = 1, 2, …, P, n = 1, 2, …, T0f s , m = 1, 2, …, M

[0018] The data is divided into P segments. τ p represents the p-th segment of the signal. The length of each segment is T0 seconds, and the segmentation step size is T b seconds. f s is the sampling rate of the signal. τ p represents slow time, n represents fast time, and m is the array element number.

[0019] A moving target detection method based on spatial slices of spatio-temporal frequency transformation. Specifically, Step 2 is as follows

[0020]

[0021] Within T0 seconds, the frequency ω p and azimuth angle α(τ p ) of the received signal are approximately invariant. is the frequency value received by the array at time τ p after being affected by the Doppler effect caused by the target movement, where is the target velocity vector, is the wave number vector; is the distance vector from the target to the central receiving array element at the initial moment, d is the array element spacing, ω0 and f0 are both the frequencies of the single-frequency signal emitted by the target.

[0022] A moving target detection method based on space-time-frequency transformation space slicing, the specific step 3 is as follows,

[0023]

[0024] In the formula, α = α(τ p ) represents the azimuth angle, M represents the number of array elements of the array, m represents the m-th array element, τ p represents the p-th segment of the signal, i represents the imaginary unit, cosα represents the cosine value of the azimuth angle, c represents the speed of sound, N represents the number of points in each segment of the signal after segmentation, N - 1 represents the point position, is the wave number vector; is the distance vector from the target to the central receiving array element at the initial moment.

[0025] A moving target detection method based on space-time-frequency transformation space slicing, the specific step 3 is as follows:

[0026] According to the space transformation mathematical expression:

[0027]

[0028] where γ0 ∈ [0, 360), γ0 is the angle measured clockwise from the true north line to the target course line;

[0029] Search for the angle γ0, and then perform a coordinate transformation on the (f - α - t) three-dimensional space according to this parameter as shown in Equation (4) to obtain a new three-dimensional space (f - cosθ - t);

[0030]

[0031] Then the signal is represented as a curve on the plane in the new three-dimensional space.

[0032] A moving target detection method based on space-time-frequency transformation space slicing, the specific step 5 is as follows: Search for the transmitted signal frequency f0 and the velocity v, and determine the slice A in the three-dimensional space according to these two parameters:

[0033] A computer-readable storage medium stores a computer program therein, and when the computer program is executed by a processor, it implements the method steps described in any one of claims 1 - 6.

[0034] An electronic device includes a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete communication with each other through the communication bus;

[0035] The memory is used to store computer programs;

[0036] When the processor is used to execute the program stored in the memory, it implements the method steps described in any one of claims 1-6.

[0037] The beneficial effects of the present invention are:

[0038] The present invention improves the detection performance of the underwater sonar array.

[0039] The method of the present invention is simple, fast, and effective.

[0040] The present invention increases the processing time of moving target signals.

[0041] The present invention can better display the changes of frequency and azimuth over time compared with the LOFAR diagram and the azimuth history diagram.

[0042] The present invention is conducive to the detection of targets with large changes in azimuth over time.

[0043] The algorithm of the present invention has a low computational complexity. Description of the Drawings

[0044] Figure 1 It is a schematic diagram of the target motion model of the present invention.

[0045] Figure 2 It is a flowchart of the method of the present invention.

[0046] Figure 3 It is a schematic diagram of the position of the signal of the present invention in three-dimensional space, Figure 3 -(a) A curve diagram representing the signal in the three-dimensional space of (f-α-t), Figure 3 -(b) A top view of the three-dimensional space of (f-α-t), Figure 3 -(c) A side view of the three-dimensional space of (f-α-t), Figure 3 -(d) A front view of the three-dimensional space of (f-α-t).

[0047] Figure 4 It is a schematic diagram of the position of the signal of the present invention in the three-dimensional transformation space, Figure 4 -(a) A curve diagram representing the signal in the three-dimensional space of (f-cosθ-t), Figure 4 -(b) A top view of the three-dimensional space of (f-cosθ-t), Figure 4 -(c) A side view of the three-dimensional space of (f-cosθ-t), Figure 4 -(d) A front view of the three-dimensional space of (f-cosθ-t).

[0048] Figure 5 It is a schematic diagram of taking slices in the three-dimensional space of the present invention.

[0049] Figure 6 It is the LOFAR diagram of the present invention, Figure 6 -(a) LOFAR diagram at 64 Hz, Figure 6 -(b) LOFAR diagram at 67 Hz, Figure 6 -(c) LOFAR diagram at 70 Hz, Figure 6 -(d) LOFAR diagram at 73 Hz.

[0050] Figure 7 It is the azimuth history diagram of the present invention, Figure 7 -(a) Azimuth history diagram at 64 Hz, Figure 7 -(b) Azimuth history diagram at 67 Hz, Figure 7 -(c) Azimuth history diagram at 70 Hz, Figure 7 -(d) Azimuth history diagram at 73 Hz.

[0051] Figure 8 It is the effect diagram of the present invention, Figure 8 -(a) Effect diagram at 64 Hz, Figure 8 -(b) Effect diagram at 67 Hz, Figure 8 -(c) Effect diagram at 70 Hz, Figure 8 -(d) Effect diagram at 73 Hz. Detailed implementation manners

[0052] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0053] A moving target detection method based on spatio-temporal-frequency transformation space slicing, the moving target detection method includes the following steps:

[0054] Step 1: Segment the target radiation sound signal s(t) received by an M-element horizontal line array in a low signal-to-noise underwater acoustic environment;

[0055] Step 2: For each period τ in Step 1 p of the received signals on each array element, perform an N-point DFT respectively, where N = T0·f s ;

[0056] The azimuth angle α(τ p ), each period τp There is an azimuth angle α(τ p ), step S3 compensates the azimuth angle for each signal segment; however, the azimuth angle is unknown during compensation, so it will be compensated once from 0 to 180 degrees. In the formula, α is a variable, α(τ p ) is a constant.

[0057] Step 3: Perform frequency domain beamforming on each segment of the array signal after DFT in step 2, and compensate each element by the azimuth angle α(τ p ) brings about the phase difference between the arrays and then superimposes;

[0058] Step 4: Perform coordinate transformation on the frequency-azimuth-time (f-α-t) three-dimensional matrix space obtained in step 3;

[0059] Step 5: Take a slice in the frequency-azimuth-time (f-cosθ-t) three-dimensional space of the coordinate transformation obtained in step 4;

[0060] Step 6: Perform segmented Radon transform on the spatial slice obtained in step 5 to detect the target.

[0061] A moving target detection method based on space slicing of time-space frequency transform, wherein the step 1 is specifically as follows:

[0062] s m (τ p ,n)=s((τ p -1)T b +n) (1)

[0063] τ p =1,2,…,P, n=1,2,…,T0f s ,m=1,2,…,M

[0064] The data is divided into P segments, τ p Represents the pth segment signal, each segment signal length is T0 seconds, and the segment step is T b Seconds, f s is the sampling rate of the signal, τ p represents slow time, n represents fast time, and m is the array element number.

[0065] A moving target detection method based on space slicing of time-space frequency transform, wherein the step 2 is specifically as follows:

[0066]

[0067] In T0 seconds, the frequency of the received signal ω p , azimuth angle α(τ p ) is approximately unchanged, is the frequency value received by the array at time τ after being affected by the Doppler effect caused by the target motion, where p is the target velocity vector, and is the wave number vector; is the distance vector between the target and the central receiving element at the initial time, d is the element spacing, and ω0 and f0 are both the frequencies of the single-frequency signal emitted by the target.

[0068] A moving target detection method based on space-time-frequency transformation space slicing. The specific steps of step 3 are as follows:

[0069]

[0070] In the formula, α = α(τ p ) represents the azimuth angle, M represents the number of array elements, m represents the m-th element, τ p represents the p-th segment of the signal, i represents the imaginary unit, cosα represents the cosine value of the azimuth angle, c represents the speed of sound, N represents the number of points in each segment of the signal after segmentation, and N - 1 represents the point position. is the wave number vector; is the distance vector between the target and the central receiving element at the initial time.

[0071] A moving target detection method based on space-time-frequency transformation space slicing. The specific steps of step 3 are as follows:

[0072] According to the space transformation mathematical expression:

[0073]

[0074] where γ0 ∈ [0, 360), γ0 is the angle measured clockwise from the true north line to the target course line and is a constant;

[0075] Search for the angle γ0, and then perform a coordinate transformation on the (f - α - t) three-dimensional space according to this parameter as shown in Equation (4) to obtain a new three-dimensional space (f - cosθ - t);

[0076]

[0077] Then the signal in the new three-dimensional space is represented as a curve on the plane

[0078] A moving target detection method based on space-time-frequency transformation space slicing. The specific steps of step 5 are as follows: Search for the transmitted signal frequency f0 and the velocity v, and determine the slice A in the three-dimensional space according to these two parameters:

[0079] ​A computer-readable storage medium stores a computer program therein, and when the computer program is executed by a processor, the method steps described above are implemented.

[0080] An electronic device includes a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete communication with each other through the communication bus;

[0081] The memory is used for storing a computer program;

[0082] The processor is used for implementing the method steps described above when executing the program stored on the memory.

[0083] Figure 1 The established target motion model is shown. First, it is assumed that the receiving device deployed underwater is a uniform horizontal linear array with M array elements. The reference element and the equivalent acoustic center of the linear array are located at the coordinate origin. The moving target starts from the initial point A(x0, y0) and moves at a speed in a uniform straight line motion, radiating a single-frequency signal with a radiation angular frequency of ω0. The course γ0, γ0 ∈ [0, 360), is the angle measured clockwise from the true north line to the target course line, and is a constant. At time t, the target moves to point B, and the beam angle θ, θ ∈ [0, 180], is the angle between the current target motion direction and the position distance vector of the target . α is the target azimuth angle, that is, the vector of the line connecting the target position and the equivalent acoustic center of the array and the included angle with the x-axis. θ and α change continuously with the target motion or the passage of time, and can be expressed as functions of time t, θ(t) and α(t). Only considering the case where the target moves in the first and second quadrants, the positional relationship among the three angles cosθ, α, and γ0 is:

[0084] cosθ(t) = -sin(α(t) + γ0)

[0085] Assume that the sound field is an infinite free plane, and the receiving array is in the far field of the signal source. Then the received signal can be approximately regarded as a plane wave. During a period of time, due to the Doppler effect caused by the target motion, the signal received by the element at the coordinate origin is:

[0086]

[0087] where, is the wave number vector. The coordinate of any element located on the x-axis can be expressed as (x, 0). According to the principle of plane wave incidence, the phase difference between the signals received by this element and the central element is Then the signal received by any element can be written as:

[0088]

[0089] From Figure 1 the moving target model, it can be seen that the azimuth angle of the target changes with time and also changes with the radial velocity of the target's movement. Due to the Doppler effect, the received frequency also changes with time, and a non-stationary signal is received. Therefore, a spatio-temporal two-dimensional short-time Fourier transform is performed on the above array signal, that is, the source signal s(t', x) is multiplied by the time window function γ*(t' - t) to achieve the interception and translation of the source signal. Then, the signal in the space-time domain is subjected to a two-dimensional Fourier transform to obtain the frequency-azimuth domain of the signal. The result of the transformation can be expressed by the following formula:

[0090]

[0091] Substituting and simplifying Equation (3) gives:

[0092]

[0093] By sliding the window function and performing a two-dimensional Fourier transform, the functional expression of the received signal in the three-dimensional space represented by the three axes of frequency-azimuth-time (f-α-t) is obtained. As Figure 3 shown, it is a curve on the surface in the three-dimensional space.

[0094] From the above derivation, it can be known that the frequency of the received signal at any time is:

[0095]

[0096] Since f0, v, and c are all constants, the relationship between f(t) and cosθ(t) is linear. Therefore, according to the mathematical expression of space transformation:

[0097]

[0098] Search for the angle γ0, and perform a coordinate transformation on the (f-α-t) three-dimensional space according to this parameter as shown in Equation (6) to obtain a new three-dimensional space (f-cosθ-t). The position of the signal in the (f-cosθ-t) is as Figure 4 shown. Since the relationship between f(t) and cosθ(t) is linear, it can be seen that Figure 4 in the top view of (b), the points representing the signal are connected into a straight line, that is, the signal in the new three-dimensional space can be represented as a curve on the plane and its equation is:

[0099]

[0100] Search for the transmitted signal frequency f0 and velocity v. According to these two parameters, the slice A in the three-dimensional space can be determined: Namely Figure 5 the plane represented by the blue block diagram in

[0101] For example Figure 8 , the data of the Swellex96 S5 experiment from 0 to 32 minutes is processed using the present invention and compared with the LOFAR diagram of Figure 6 and the azimuth history diagram of Figure 7 . The left figure is the slice taken in space, and the right figure is the result after performing the Radon transform on the slice.

Claims

1. A moving target detection method based on spatio-temporal-frequency transformed spatial slices, characterized in that The method for detecting a moving target includes the following steps: Step 1: Segment the target radiated acoustic signal received by the element horizontal line array in a low signal-to-noise ratio underwater acoustic environment ; Specifically, step 1 is as follows: When the data is divided into segments, indicating the segment signal, and the length of each segment signal is seconds, with a segmentation step size of seconds, being the sampling rate of the signal, indicating the slow time, indicating the fast time, being the array element serial number; Step 2: For each period in Step 1 perform N-point DFT on the received signals at each array element respectively, ; Step 3: Perform frequency-domain beamforming on each piece of the array signal after DFT in Step 2, and superimpose them after compensating the phase difference between the array elements caused by the azimuth angle ; Step 4: Perform coordinate transformation on the frequency-azimuth-time three-dimensional matrix space obtained in Step 3; Step 5: Take a slice in the three-dimensional space of the frequency-azimuth-time of the coordinate transformation obtained in Step 4 ; Step 6: Perform a segmented Radon transform on the spatial slice obtained in step 5 to detect the target.

2. The method for detecting a moving target based on a space slice of spatio-temporal-frequency transformation according to claim 1, wherein Specifically, step 2 is as follows: （2） Within seconds, the frequency and azimuth angle of the received signal are approximately invariant. is the frequency value received by the array at time after being affected by the Doppler effect caused by the target motion, where is the target velocity vector, is the wave number vector; is the distance vector between the target and the central receiving element at the initial time, is the element spacing, are all the frequencies of the single-frequency signal emitted by the target.

3. The method for detecting a moving target based on a space-time-frequency transformed space slice according to claim 2, wherein Specifically, step 3 is as follows: （3） In the formula, = represents the azimuth angle, M represents the number of array elements of the array, represents the m-th array element, represents the segment signal, represents the imaginary unit, represents the cosine value of the azimuth angle, represents the speed of sound, represents the number of points in each segment of the signal after segmentation, represents the point position, is the wave number vector; is the distance vector between the target and the central receiving array element at the initial moment.

4. The method for detecting a moving target based on a spatial slice of spatio-temporal frequency transformation according to claim 1, wherein Specifically, step 3 is: According to the mathematical expression of spatial transformation: （4） Among them , is the angle measured clockwise from the true north line to the target course line; Search angle , then perform coordinate transformation on the three-dimensional space as shown in Equation (4) according to this parameter to obtain a new three-dimensional space ; ; The signal is then represented as a curve on the plane in the new three-dimensional space.

5. The method for detecting a moving target based on a spatial slice of spatio-temporal-frequency transformation according to claim 1, wherein The specific content of step 5 is to search for the transmitting signal frequency and speed , and determine the slice in three-dimensional space based on these two parameters : .

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method steps described in any one of claims 1-5 are implemented.

7. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete communication with each other through the communication bus; The memory is used to store a computer program; The processor is used to implement the method steps described in any one of claims 1-5 when executing the program stored on the memory.

Images