Grating lobe suppression active detection method and device based on frequency differential beamforming

Abstract

The present invention relates to a grating lobe suppression active detection method and device based on frequency differential beamforming. The method combines the matched filter algorithm commonly used in active detection with the frequency differential beamforming algorithm. First, pulse compression is performed through the matched filter, and then the spatial directional target direction is formed by the frequency differential beamforming algorithm. Finally, the active detection display result is obtained through downsampling and smoothing. The suspected target position is manually marked to obtain the direction and distance information. The device includes a sound source transmitter, a linear array sonar, a bandpass filter, a matched filter, a signal processor, and a sonar display and control. The sound source transmits a detection signal; the linear array sonar receives the reflected signal; the bandpass filter, the matched filter, and the signal processor process the signal to form a waterfall diagram in the sonar display and control to obtain the direction and distance of the detected target. The present invention solves the problem of grating lobe generation in underwater active detection and meets the real-time and stability requirements in actual engineering.

Description

Technical Field

[0001] The present invention belongs to the field of underwater active detection, and in particular relates to a grating lobe suppression active detection method and device based on frequency differential beamforming, which can be applied to various linear array sonar systems to solve the grating lobe problem caused by the mismatch between the detection signal and the sonar receiving system frequency band. Background Art

[0002] This invention is based on the demand for sonar active detection technology. To increase the detection range, the sonar array needs to adopt a larger element spacing during design. A larger element spacing often means that the transmitting sound source needs to use a lower frequency detection signal, which is often not met. When the transmitting frequency does not match the array operating frequency, a serious grating lobe effect will occur. This invention proposes a matched filtering method based on frequency differential beamforming, which can effectively suppress the generation of grating lobes and meet real-time computing requirements. This method has been verified in actual engineering projects. Summary of the Invention

[0003] The present invention aims to overcome the serious impact of grating lobes in current sonar active detection technology. It describes a grating lobe suppression active detection method and device based on frequency differential beamforming. The method combines the matched filter, a commonly used algorithm in active detection, with the frequency differential beamforming (FDB) algorithm. First, pulse compression is performed using the matched filter. Then, the frequency differential beamforming algorithm generates spatial directivity to determine the target's direction. Finally, downsampling and smoothing are used to determine the target's direction and range.

[0004] The present invention proposes a grating lobe suppression active detection method based on frequency differential beamforming. This method combines the matched filtering method commonly used in active detection with the frequency differential beamforming algorithm. First, pulse compression is performed through a matched filter; then, the frequency differential beamforming algorithm is used to form spatial directivity to determine the target direction; finally, the active detection display result is obtained through downsampling and smoothing, suspected targets are marked, and the direction and distance information of the detected target are obtained.

[0005] As an improvement to the above technical solution, the present invention proposes an active detection grating lobe suppression method based on frequency differential beamforming, the method comprising the following steps:

[0006] 1) First, a detection signal is emitted from an acoustic source. The signal is reflected by the target and the linear array sonar receives the original signal data reflected by the target.

[0007] 2) Secondly, the raw signal data received by the linear array sonar is divided into multiple channels according to the number of array elements of the linear array sonar and transmitted to each array element of the linear array sonar. The signal data of each channel is subjected to bandpass filtering and matched filtering to determine the peak point of the target on the single array element domain signal within a single ping-pang period. The ping-pang period is the time from the sound source transmitting the signal to receiving the target's reflected signal;

[0008] 3) Then, frequency differential beamforming is performed on the array signal data after matched filtering to suppress the generation of grating lobes and obtain a spatial image containing the target orientation;

[0009] 4) Finally, by extracting and smoothing the sampling points, the specific position and distance of the detection target are obtained in the space image.

[0010] As another improvement of the above technical solution, in step 1), the sound source uses linear frequency modulation technology LFM to transmit the detection signal.

[0011] As another improvement to the above technical solution, after the linear array sonar obtains the time domain signal of each array element, the method takes data of a certain sampling point length from the time domain signal of each array element to form a data snapshot as the data input of each channel, which is expressed as:

[0012] X(t)=[x1(k) x2(k) ... x M (k)] T

[0013] Among them, X(t) is the data snapshot, M is the number of array elements set, and x i (k) is the received signal of the i-th array element, x i (k) = [x i (1) x i (2) ... x i (L)], where L is the sampling point length of the data snapshot.

[0014] As another improvement of the above technical solution, when performing matched filtering on the signal data, the method receives the signal x for each array element. i (k) performs sliding cross-correlation processing with the copy signal s(t), where the copy signal s(t) is a replica of the transmitted signal. The expression of the output result after processing is:

[0015]

[0016] Where T is the copy signal length, t is time, p i (t) Output the results after matched filtering for each channel.

[0017] As another improvement to the above technical solution, the method performs frequency differential beamforming processing on the signal data of each channel, and the output result expression after processing is:

[0018]

[0019] in, is the beam output result, is the average frequency of the signal, Δf is the frequency difference, θ is the incident direction of the signal, f L is the lower limit frequency of the transmitted signal, f H is the upper frequency limit of the transmitted signal, For conduct Calculates the average value within a defined range.

[0020] As a further improvement to the above technical solution, the method extracts and smoothes sampling points from the output results of frequency differential beamforming. Specifically, the method comprises: dividing the spatial image signal data output by the frequency differential beamforming into multiple directions, extracting and smoothing the signal data from each direction, and forming a waterfall chart display within a ping-pang period;

[0021] According to the results displayed in the waterfall chart, the direction and distance of the detected target are obtained by marking.

[0022] The present invention proposes an active detection device for grating lobe suppression based on frequency differential beamforming based on one of the above methods, the device comprising a sound source transmitter, a linear array sonar, a bandpass filter, a matched filter, a signal processor, and a sonar display and control;

[0023] The sound source transmitter is used to transmit the original detection signal;

[0024] The linear array sonar is used to receive signals reflected by the detected target and divide the received signals into multiple channels as raw data;

[0025] The bandpass filter is used to perform bandpass filtering on the signal of each channel received by the linear array sonar;

[0026] The matched filter is used to perform pulse compression processing on the signal of each channel after being processed by the bandpass filter;

[0027] The signal processor is used to perform differential beamforming on the signal of each channel after being processed by the matched filter, and form a spatial image by a frequency differential beamforming algorithm; then, sampling points are extracted and smoothed in multiple directions in the spatial image in sequence, and the processed results are transmitted to the sonar display and control;

[0028] The sonar display and control is used to form a waterfall diagram of the received results, and the direction and distance of the detected target are obtained by marking in the waterfall diagram.

[0029] As an improvement of the above technical solution, the sound source transmitter is a fish-lip transducer.

[0030] As another improvement of the above technical solution, the linear array sonar is composed of multiple hydrophones arranged in a towed array or a broadside array.

[0031] The advantages of this invention are that it effectively solves the problem of grating lobes caused by the mismatch between the transmission frequency and the array element spacing during active detection, while also meeting real-time and stability requirements. Experimental data from the application examples demonstrates the effectiveness of the method. The invention can be applied to various linear array sonar active detection systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a flow chart of the overall algorithm of the detection method of the present invention;

[0033] Figure 2 This is a system flow chart of the detection method of the present invention;

[0034] FIG3 is a comparison diagram of simulation results of a conventional method and the method of the present invention, wherein FIG3(a) is a result diagram of the conventional detection method, FIG3(b) is a diagram of the instantaneous azimuth amplitude at the distance of the target marked in (a), FIG3(c) is a diagram of the result diagram of the detection method of the present invention, and FIG3(d) is a diagram of the instantaneous azimuth amplitude at the distance of the target marked in (c);

[0035] Figure 4 is a comparison of the data processing results of a sea trial using the conventional method and the method of the present invention, wherein Figure 4(a) is a result diagram of the conventional detection method, Figure 4(b) is a diagram of the instantaneous azimuth amplitude at the target distance marked in (a), Figure 4(c) is a diagram of the result of the detection method of the present invention, and Figure 4(d) is a diagram of the instantaneous azimuth amplitude at the target distance marked in (c). DETAILED DESCRIPTION

[0036] The specific implementation of the present invention is further described in detail below with reference to data from a certain offshore test and the accompanying drawings.

[0037] The detailed process adopted by the present invention is as follows Figure 1 As shown in the figure, first, bandpass filtering and matched filtering are performed on each channel based on the raw array data received by the linear array sonar to determine the peak point of the target within a single ping-pang cycle; then, frequency differential beamforming is performed on the array data after pulse compression, which can effectively suppress the generation of grating lobes while obtaining the target direction; finally, the direction and distance of the detected target can be obtained by extracting and smoothing the sampling points.

[0038] The specific steps are described as follows:

[0039] (1) Linear array data input

[0040] The linear array sonar receives spatial signals and obtains time domain signals of M array elements. Data with a length of L sampling points is taken to form a data snapshot as data input:

[0041] X(t)=[x1(k) x2(k) ... x M (k)] T

[0042] Among them, x i (k) = [x i (1) x i (2) ... x i (L)];

[0043] (2) Matched filter

[0044] For each array element, receive signal x i (k) is subjected to sliding cross-correlation processing with the copy signal s(t), and the output result can be expressed as follows:

[0045]

[0046] Where T is the copy signal length, p i (t) Output the results after matched filtering for each channel.

[0047] (3) Differential beamforming

[0048] After matched filtering, the time domain signal received by each array element is transformed by FFT, and the frequency domain expression is as follows:

[0049]

[0050] Where f is the signal frequency, A i (f) is the spectrum of the reflected signal of the i-th target, d is the distance between adjacent array elements, c is the speed of sound, θ i is the incoming wave direction of the i-th target.

[0051] Defined as and The product of is expressed as follows:

[0052]

[0053] Where, the superscript * represents the complex conjugate. Δf represents the frequency difference, satisfying the inequality 1 / T<Δf<f H -f LAt the same time, in order to suppress the generation of side lobes, the inequality Δf≤c / (2d) must be satisfied. f represents the frequency to be detected, satisfying the inequality f L ≤f≤f H -Δf,f L With f H The upper and lower limit frequencies of the detection signal are transmitted respectively.

[0054] The steering vector for differential beamforming is shown in the following equation:

[0055] w m (Δf,sinθ)=exp(-j2πΔf(m-1)dsinθ / c)

[0056] frequency The relationship between the complex beam output at and the frequency difference Δf is shown in the following equation:

[0057]

[0058] Combining the above equations, we can get:

[0059]

[0060] For a given Δf, the broadband beam output of FDB can be obtained in an incoherent manner The average value of is the final output of frequency differential beamforming:

[0061]

[0062] According to the spatial sampling theorem, grating lobes are caused by d / lambd > 1 / 2, where d is the array element spacing and lambd is the wavelength. Reducing the array element spacing or lowering the source frequency can suppress grating lobes. However, when the array element spacing and the source frequency are fixed, this frequency difference can be used to reduce the frequency and thus avoid grating lobes.

[0063] (4) The sampling points of the output results of the frequency differential beamforming are extracted and smoothed to form a waterfall diagram within the ping-pang period to obtain the direction and distance of the detected target.

[0064] Downsampling is done by passing the signal through a low-pass filter with a bandwidth of F / (2M), where F is the original signal bandwidth and M is the sampling interval. After passing through the filter, the signal is sampled a second time with a decimation interval of M.

[0065] The input data used in this example is from marine tests of a linear array sonar. The array has 16 hydrophones (M), with a spacing of 7.5 meters (d), an array length of 120 meters, a sampling rate of 128 kHz, and a sound velocity of 1450 meters per second (c). The active transmit source has a bandwidth of B = 600-800 Hz, a pulse width of 0.5 seconds (tao), a pulse repetition period of 60 seconds (T), and a beamforming interval of 1°, providing a beamforming range of 1 to 180°.

[0066] like Figure 1 FIG. 1 is a flow chart of the overall algorithm of the detection method of the present invention;

[0067] The present invention proposes a grating lobe suppression active detection method based on frequency differential beamforming, and the specific steps are as follows:

[0068] Step 1: Correspondence Figure 1 In step 101, for the sampling signal of each array element in M=16 array elements, array data with a length of L=tao*fs sampling points is taken to form a snapshot as data input.

[0069] X(t)=[x1(k) x2(k) ... x M (k)] T

[0070] Among them, x i (k) is the received signal of the i-th array element, x i (k) = [x i (1) x i (2) ... x i (L)], where L is the sampling point length of the data snapshot.

[0071] Step 2: Correspond Figure 1 In step 102, time domain matched filtering is performed on the data snapshot X(t), and the specific process is as follows.

[0072]

[0073] Where T is the copy signal length, p i (t) Output the results after matched filtering for each channel.

[0074] Step 3: Correspond Figure 1 Step 103 in the example, after matched filtering, the output result p of each element after matched filtering is i (t) Perform FFT transformation, and the frequency domain expression can make the following model assumptions:

[0075]

[0076] Where f is the signal frequency, A i(f) is the spectrum of the reflected signal of the i-th target, m is the array element number, d is the distance between adjacent array elements, c is the speed of sound, θ i is the incoming wave direction of the i-th target.

[0077] For the i-th target, the m-th array element is The value at the frequency point, For the i-th target, the m-th array element is The value at Defined as and The product of is expressed as follows:

[0078]

[0079] Where, the superscript * represents the complex conjugate. Δf represents the frequency difference, satisfying the inequality 1 / T<Δf<f H -f L At the same time, in order to suppress the generation of side lobes, the inequality Δf≤c / (2d) must be satisfied. f represents the frequency to be detected, satisfying the inequality f L ≤f≤f H -Δf,f L With f H The upper and lower limit frequencies of the detection signal are transmitted respectively.

[0080] The steering vector for differential beamforming is shown in the following equation:

[0081] w m (Δf,sinθ)=exp(-j2πΔf(m-1)dsinθ / c)

[0082] frequency The complex beam output at The relationship with the frequency difference Δf is shown as follows:

[0083]

[0084] Combining the above equations, we can get:

[0085]

[0086] For a given Δf, the broadband beam output of FDB can be obtained in an incoherent manner The average value of , that is, the final output result of frequency differential beamforming is expressed as:

[0087]

[0088] Here, the spectrum signal is modeled, assuming that the phase shift of the i-th target signal spectrum is as follows. In practice, the matched filtering result can be directly processed using the FFT as input for this algorithm. While the i-th target is assumed during the derivation process, in practice, a full 180-degree scan is performed, assuming that targets appear at different locations.

[0089] Step 4: Correspondence Figure 1 In steps 104 and 105, sampling points are extracted and smoothed on the output results of the frequency differential beamforming to form a waterfall diagram within the ping-pang period.

[0090] Step 5: Correspondence Figure 1 In step 106, the direction and distance of the detected target are obtained.

[0091] like Figure 2 FIG. 1 is a system flow chart of the detection method of the present invention, which specifically includes the following steps: first, the raw array data received by the linear array sonar is divided into N channels, where N is a set number, and each channel is subjected to bandpass filtering and matched filtering to determine the peak point of the target within a single ping-pang cycle; second, frequency differential beamforming is performed on the pulse-compressed array data to obtain a spatial image containing the target's orientation while effectively suppressing the generation of grating lobes; finally, the spatial image containing the target's orientation is extracted and smoothed according to the N orientations, and the orientation and distance of the detected target are obtained in the sonar display and control device.

[0092] Conventional methods use time-domain beamforming or frequency-domain beamforming methods, then perform matched filtering on the beam output results, and finally perform subsequent processing such as downsampling and smoothing.

[0093] As shown in Figure 3, it is a comparison diagram of the simulation results of the conventional method and the method of the present invention, wherein Figure 3(a) is the result diagram of the conventional detection method, Figure 3(b) is the instantaneous azimuth amplitude diagram at the target distance marked in (a), Figure 3(c) is the result diagram of the detection method of the present invention, and Figure 3(d) is the instantaneous azimuth amplitude diagram at the target distance marked in (c);

[0094] In Figures 3(a) and 3(b), due to the influence of grating lobes, multiple false targets appear at the same distance, resulting in a high false alarm rate. In Figures 3(c) and 3(d), the grating lobes are well suppressed, the false alarm rate is low, and true targets can be easily marked manually or automatically.

[0095] As shown in Figure 4, the comparison results of the data processing of a certain sea trial using the conventional method and the method of the present invention are shown, wherein Figure 4(a) is the result diagram of the conventional detection method, Figure 4(b) is the instantaneous azimuth amplitude diagram at the target distance marked in (a), Figure 4(c) is the result diagram of the detection method of the present invention, and Figure 4(d) is the instantaneous azimuth amplitude diagram at the target distance marked in (c);

[0096] In Figures 4(a) and 4(b), due to the influence of grating lobes, multiple false targets appear at the same distance, making it impossible to find the true target. In Figures 4(c) and 4(d), the grating lobes are well suppressed, the target false alarm rate is significantly reduced, and the true target can be easily marked manually or automatically.

[0097] It can be seen from the above detailed description of the present invention that the present invention can achieve accurate detection of the target position and direction while effectively suppressing the grating lobe.

[0098] Finally, it should be noted that the above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention and are intended to be encompassed by the claims of the present invention.

Claims

1. A grating lobe suppression active detection method based on frequency differential beamforming. This method combines the matched filtering method commonly used in active detection with the frequency differential beamforming algorithm. First, pulse compression is performed using a matched filter. Then the frequency differential beamforming algorithm is used to form spatial directivity to determine the target direction; Finally, the active detection display results are obtained through downsampling and smoothing, the suspected targets are marked, and the direction and distance information of the detected targets are obtained; The method comprises the following steps: 1) First, a detection signal is emitted from an acoustic source. The signal is reflected by the target and the linear array sonar receives the original signal data reflected by the target. 2) Secondly, the raw signal data received by the linear array sonar is divided into multiple channels according to the number of array elements of the linear array sonar and transmitted to each array element of the linear array sonar. The signal data of each channel is subjected to bandpass filtering and matched filtering to determine the peak point of the target on the single array element domain signal within a single ping-pang period. The ping-pang period is the time from the sound source transmitting the signal to receiving the target's reflected signal; 3) Then, frequency differential beamforming is performed on the array signal data after matched filtering to suppress the generation of grating lobes and obtain a spatial image containing the target orientation; 4) Finally, by extracting and smoothing the sampling points, the specific position and distance of the detected target are obtained in the spatial image; The method performs frequency differential beamforming processing on the signal data of each channel, and the output result expression after processing is: in, is the beam output result, is the average frequency of the signal, Δf represents the frequency difference, satisfying the inequality 1 / T<Δf<f H -f L At the same time, in order to suppress the generation of side lobes, the inequality Δf≤c / (2d) must be satisfied. For a given Δf, the broadband beam output of frequency differential beamforming can obtain the signal average frequency in an incoherent way. d is the distance between adjacent array elements, θ is the signal incident direction, f L is the lower limit frequency of the transmitted signal, f H is the upper frequency limit of the transmitted signal, For conduct Calculates the average value within a defined range.

2. The grating lobe suppression active detection method based on frequency differential beamforming according to claim 1, characterized in that: In the step 1), the sound source transmits the detection signal using linear frequency modulation (LFM) technology.

3. The active detection grating lobe suppression method based on frequency differential beamforming according to claim 1, characterized in that: After the linear array sonar obtains the time domain signal of each array element, the method takes data of a certain sampling point length from the time domain signal of each array element to form a data snapshot as the data input of each channel, which is expressed as: X(t)=[x1(k)x2(k)...x M (k)] T Among them, X(t) is the data snapshot, M is the number of array elements set, and x i (k) is the received signal of the i-th array element, x i (k) = [x i (1)x i (2)...x i (L)], where L is the sampling point length of the data snapshot.

4. The active detection grating lobe suppression method based on frequency differential beamforming according to claim 1, characterized in that: When the method performs matched filtering on the signal data, each array element receives the signal x i (k) Perform sliding cross-correlation processing with the copy signal s(t), where the copy signal s(t) is a copy of the transmitted signal. The output result after processing is expressed as: Where T is the copy signal length, t is time, p i (t) Output the results after matched filtering for each channel.

5. The grating lobe suppression active detection method based on frequency differential beamforming according to claim 1, characterized in that: The method extracts and smoothes sampling points from the output of frequency differential beamforming. Specifically, the method comprises the following steps: dividing the spatial image signal data output by the frequency differential beamforming into multiple directions, extracting and smoothing the signal data from each direction, and forming a waterfall chart display within a ping-pang period; According to the results displayed in the waterfall chart, the direction and distance of the detected target are obtained by marking.

6. A grating lobe suppression active detection device based on frequency differential beamforming according to the method of any one of claims 1 to 5, characterized in that: The device includes a sound source transmitter, a linear array sonar, a bandpass filter, a matching filter, a signal processor, and a sonar display and control; The sound source transmitter is used to transmit the original detection signal; The linear array sonar is used to receive signals reflected by the detected target and divide the received signals into multiple channels as raw data; The bandpass filter is used to perform bandpass filtering on the signal of each channel received by the linear array sonar; The matched filter is used to perform pulse compression processing on the signal of each channel after being processed by the bandpass filter; The signal processor is used to perform differential beamforming on the signal of each channel after being processed by the matched filter, and form a spatial image by a frequency differential beamforming algorithm; then, sampling points are extracted and smoothed in multiple directions in the spatial image in sequence, and the processed results are transmitted to the sonar display and control; The sonar display and control is used to form a waterfall diagram of the received results, and the direction and distance of the detected target are obtained by marking in the waterfall diagram.

7. The grating lobe suppression active detection device based on frequency differential beamforming according to claim 6, characterized in that: The sound source transmitter is a fish-lip transducer.

8. The grating lobe suppression active detection device based on frequency differential beamforming according to claim 6, characterized in that: The linear array sonar is composed of a plurality of hydrophones arranged in a towed array or a broadside array.

Images