A beam domain signal simulation method based on array horizontal beam characteristics pre-computation

Abstract

The application discloses a kind of based on array horizontal beam characteristic precomputation's beam domain signal simulation method, it is related to technical field;Its technical points are: by being carried out once in array horizontal beam characteristic simulation before calculation, and with corresponding equipment binding storage, subsequent all using the simulation scene of equipment can directly call precomputation result, without repeating matrix operation;Single target simulation only needs to calculate the sound field between reference array element and target, without traversing all array elements to carry out independent sound field calculation, so that simulation operation amount no longer increases with the increase of array element number, can significantly shorten the signal simulation time length of large aperture array, while greatly reducing the dependence on computer hardware computing power and memory capacity, can complete the signal simulation task of large-scale array on portable simulation equipment.

Description

Technical Field

[0001] This invention relates to the field of sonar beam domain signal simulation technology, specifically a beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics. Background Technology

[0002] Sonar signal simulation is a core technology that uses signal simulation algorithms, underwater acoustic environment models, and platform equipment models to digitally simulate the entire workflow of sonar system signal source generation, channel synthesis, and array reception. It can generate high-fidelity simulated underwater acoustic signal data and is an essential support for sonar equipment algorithm verification, combat simulation, simulation teaching, and professional personnel training.

[0003] Current sonar signal simulation generally adopts a technical framework from situational assessment, source generation, channel synthesis to array signal generation. Among them, array signal generation is the core link connecting channel characteristics and array reception. Currently, two main technical approaches are used: The first is the reference element delay method, which mainly calculates the received signal of a single reference element and then applies a fixed time delay difference to the reference signal according to the geometric positional relationship between the elements, thereby simulating the received signal of all elements. Although this method is simple to calculate, it can only simulate the element time delay difference under plane wave incidence and cannot accurately reflect the spatial distribution characteristics of array directivity and multipath effects. The simulation accuracy is difficult to meet engineering requirements. The second type is the element-by-element sound field calculation method, which calculates the sound propagation sound field between each element in the array and the target independently, completes the channel synthesis and signal convolution operation for each element, and finally splices them together to obtain the complete array received signal. Although this method can ensure simulation accuracy, each additional element requires an extra complete sound field calculation, channel synthesis and signal convolution operation. As modern sonar develops towards large aperture and multiple elements, the number of array elements has increased from dozens to thousands. The computational load of element-by-element calculation has increased exponentially, resulting in a single frame signal simulation time of up to hundreds of seconds, which cannot support the simulation needs of large-scale multi-target scenarios.

[0004] Furthermore, the signals generated by existing array element domain simulation methods cannot be directly used in subsequent sonar signal processing. Since core signal processing algorithms such as moving target element calculation, spectral analysis, and matched filtering are all based on beam domain data, an additional beamforming operation must be performed on the generated array element domain signal to convert the time delay information into azimuth information. This process further increases the computation time of the entire process. To address this, we provide a beam domain signal simulation method based on the pre-calculation of array horizontal beam characteristics. Summary of the Invention

[0005] To achieve the above objectives, the present invention provides the following technical solution: A beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics, the method includes the following steps: S1. Configure the sonar equipment's operating frequency band, initial position of array elements, number of beamforming angles, and number of high-precision receiving angles. Calculate the array's horizontal beam characteristics and bind the array's horizontal beam characteristics to the corresponding sonar equipment and store them in the equipment database. S2. Construct a marine simulation scenario and obtain simulation equipment platform parameters, gridded marine environment parameters of the target sea area, and marine simulation situation information including platform heading, platform speed, number of targets, target heading, target speed, relative distance, and relative bearing; S3. Generate active transmission signals based on the active operating mode of the sonar, and generate passive platform radiated noise source signals based on the passive operating mode of the sonar. S4. Based on the gridded marine environmental parameters of the target sea area, the sound pressure result and the sound ray arrival structure are calculated using the sound propagation model. The sound ray arrival structure is then used to synthesize a broadband simulation channel that reflects the propagation loss and multipath effect of the underwater acoustic channel. S5. Determine the position of the reference array element. For a single target, only the sound field between the reference array element and the target is calculated. Based on the marine simulation situation information and equipment platform parameters, add delay effect and Doppler frequency shift effect to the active transmission signal or passive platform radiated noise source signal. Perform convolution operation on the source signal after superposition effect with the broadband simulation channel to obtain the time domain reference signal of the reference array element. S6. Based on the time-domain reference signal of the reference array element and the pre-calculated and stored array horizontal beam characteristics, the beam-domain time-domain signal is obtained through frequency-domain directivity weighting.

[0006] Furthermore, the pre-calculation of the array horizontal beam characteristics in step S1 specifically includes: S11. Calculate the array manifold matrix based on the sonar equipment parameters, wherein the column vectors of the array manifold matrix are the response vectors of the array to different beamforming angles; S12. Calculate the product of the array manifold matrix corresponding to the number of beamforming angles and the array manifold matrix corresponding to the number of high-precision receiving angles to obtain the beam characteristic matrix of the narrowband signal. S13. For broadband signals, select multiple frequency points according to a preset octave interval, calculate the beam characteristic matrix corresponding to each frequency point, and splice them together to obtain the broadband array beam characteristic matrix. S14. For multi-layer arrays, calculate the horizontal beam characteristics of each layer based on its initial position, and obtain the total horizontal beam characteristics of the multi-layer array by energy superposition.

[0007] Furthermore, in step S11, for a uniformly distributed linear array containing M elements, the angle formed by the array with respect to the k-th beam is... The response vector is: In the formula, d is the element spacing, λ is the wavelength of the acoustic signal, and the superscript T indicates the matrix transpose operation.

[0008] Furthermore, in step S13, frequency points are selected at intervals of 0.1 octaves of the lower limit frequency, and the frequency point selection density is increased for the low-frequency bands in the target line spectrum feature set; the total number of frequency points is dynamically adjusted according to the simulation accuracy requirements and the computing power of the computer hardware.

[0009] Furthermore, in step S1, the number of beamforming angles is set according to the actual number of beams and corresponding angles configured in the sonar equipment, or customized according to the user's simulation requirements; the number of high-precision receiving angles is set in 0.1° increments, or the increment precision is adjusted according to the azimuth resolution requirements and hardware level.

[0010] Furthermore, in step S3, the actively transmitted signal is any one of a single-frequency pulse signal, a linear frequency modulated signal, or a hyperbolic frequency modulated signal; the passive platform radiated noise source is generated using a general ship radiated noise simulation model, specifically: Based on the target's physical dimensions, tonnage, and speed parameters, the continuous spectrum and line spectrum levels of the platform's radiated noise are simulated and calculated. The platform's radiated noise time-domain signal, which includes both continuous spectrum and line spectrum features, is obtained by fitting the continuous spectrum and line spectrum levels.

[0011] Furthermore, in step S4, the sound propagation model adopts a ray model, and the broadband simulation channel is synthesized by time-domain superposition of the sound rays arriving at the structure.

[0012] Furthermore, in step S5, the formula for calculating the time-domain reference signal of the reference array element is as follows: In the formula, For the reference array element time-domain signal, conv is the convolution operation. The source signal is the result of superimposing the time delay effect and the Doppler frequency shift effect. The result is the underwater acoustic channel synthesis from the reference array element to the target.

[0013] Furthermore, the beam domain signal synthesis in step S6 specifically includes: S61. Perform a fast Fourier transform on the time-domain reference signal of the reference array element to obtain the frequency-domain reference signal, and then expand the frequency-domain reference signal into a dimension-matched beam frequency-domain reference signal. S62. Based on the actual azimuth of the detected target, extract the target azimuth beam characteristic matrix corresponding to the receiving angle from the pre-calculated and stored broadband array beam characteristic matrix. S63. Perform linear interpolation on the target azimuth beam characteristic matrix to obtain a continuous beam characteristic matrix with a step of 0.1Hz within the array's operating frequency band. S64. Multiply the beam frequency domain reference signal with the interpolated continuous beam characteristic matrix point by point in the frequency domain to obtain the beam frequency domain signal. S65. Perform a fast inverse Fourier transform on the beam frequency domain signal to obtain the final beam time domain signal.

[0014] Furthermore, in step S61, the formula for calculating the beam frequency domain reference signal is as follows: In the formula, The frequency domain reference signal is denoted by , FFT is the Fast Fourier Transform operation, ones(N,1) is an N-row, 1-column matrix of all 1s, N is the number of beams, ⊗ is the Kronecker product operation, and the superscript T indicates the matrix transpose operation.

[0015] This invention provides a beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics, which has the following advantages: By performing a one-time calculation before simulating the horizontal beam characteristics of the array and binding it to the corresponding equipment for storage, all subsequent simulation scenarios using the equipment can directly call the pre-calculated results without repeating matrix operations. When simulating a single target, only the sound field between the reference array element and the target needs to be calculated, without traversing all array elements for independent sound field calculation. This makes the simulation computation load no longer increase with the number of array elements, which can significantly shorten the signal simulation time of large aperture arrays and greatly reduce the dependence on computer hardware computing power and memory capacity. It can complete the signal simulation task of large-scale arrays on portable simulation equipment. Secondly, it can directly generate beam-domain and time-domain signals without requiring additional beamforming calculations after simulation. These signals can then be directly used in subsequent signal processing steps such as moving target element calculation, spectral analysis, and matched filtering, simplifying the entire sonar signal simulation and processing process and further reducing overall computation time. Furthermore, by using high-precision receiving angles in 0.1° steps for beam characteristic calculation and increasing the frequency selection density in the low-frequency bands where target line spectrum features are concentrated, it can accurately simulate the directivity characteristics of the array and the frequency selectivity of the underwater acoustic channel, ensuring the accuracy of the simulation results. This ensures that the spectral characteristics, azimuth information, and Doppler effect of the beam-domain received signal are consistent with the preset ocean simulation situation. Attached Figure Description

[0016] Figure 1 A framework diagram for beam domain signal simulation based on pre-calculated array horizontal beam characteristics; Figure 2 This is a schematic diagram of the array's three-dimensional structure and top view. Figure 3A schematic diagram of beam characteristics at a specified angle or frequency; Figure 4 Calculate the frequency distribution diagram for the horizontal beam characteristics; Figure 5 This is a situational diagram for a passive single-target simulation case. Figure 6 A comparison of the simulated power spectrum of the line spectrum continuous energy level and the power spectrum of the fitted signal; Figure 7 To fit the signal power spectrum and compare it with the LOFAR spectrum of the beam domain received signal; Figure 8 This is a single-target frequency azimuth map; Figure 9 This is a single-target orientation history diagram; Figure 10 This is a situational diagram for a passive multi-target simulation case. Figure 11 A multi-target frequency azimuth map; Figure 12 Multi-target orientation timeline; Figure 13 This is a situational diagram for an active single-target simulation case. Figure 14 Time-frequency analysis diagram of CW beam domain echo signal; Figure 15 Time-frequency analysis diagram of LFM beam domain echo signal; Figure 16 This is a azimuth distance diagram of the LFM beam domain echo signal. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] I. Array Horizontal Beam Characteristics Analysis and Pre-calculation: The response vectors of the array to incident signals in K directions are used as the array manifold matrix. If a column is given, then the response vector matrix of the narrowband signal can be represented as an array manifold matrix. Its column vector Composed of the response vector of the array to the beamforming angle, it mainly reflects the influence of the array configuration on the received signal at different angles. Essentially, it is the modulation effect of the array geometry on the phase of the incident signal. The array manifold matrix of different configurations has significant differences, which directly determines the directional characteristics of the array.

[0019] Taking a uniformly distributed linear array (ULA) as an example, the response vector of an array containing M elements to N directions has the following form: In the formula, d represents the element spacing, λ represents the wavelength of the acoustic signal, and the superscript T indicates matrix transpose operation, reflecting the basic law that the phase difference of the received signal of each element in a linear array is proportional to the sine value of the incident angle; for the three-layer circular array used, the calculation method needs to be adjusted according to the geometric characteristics of the circular array. Based on the coordinate values ​​of each element in the polar coordinate system, the phase difference of each element relative to the array center is calculated and combined to form the response vector. The response vector of the circular array does not have the sine law of the linear array, but is directly related to the circumferential angle position of the element.

[0020] If we only calculate the array response vector for a finite number of beamforming angles, then the array manifold matrix is ​​a Vandermonde determinant, as shown below: ; in, The number of array elements. The beamforming angles have favorable mathematical properties, facilitating subsequent matrix operations. For the same array, by calculating the response matrix for the beamforming angles and the response matrix for the high-precision receiving angles, and multiplying the two matrices, the beam characteristic matrix of the array can be obtained, as shown below: ; in, The array manifold corresponding to the beamforming angle has a matrix dimension of M rows and N columns, where M is the number of array elements and N is the number of beams; The array manifold corresponding to the high-precision receiving angle has a matrix dimension of M rows and K columns, where K is the number of high-precision receiving angles and L is the number of high-precision receiving angles. For example, in this calculation method, the number of beamforming angles K=180 and L=3600, that is, it is divided into 180 beams for calculation, and the beamforming angles are 0, 2, 4, ..., 360, and the high-precision receiving angles are 0, 0.1, 0.2, ..., 360.

[0021] For broadband signals, to improve computational efficiency, frequency points are extracted from 0.1 octaves of the lower limit frequency, and the beam characteristic matrix of the corresponding frequency points is calculated to obtain the broadband array beam characteristic matrix. If there are Nf frequency points to be calculated, the broadband array beam characteristic can be expressed as follows: This is an N x K x Nf three-dimensional matrix, which contains information in three dimensions: angle, beam, and frequency, and can completely describe the broadband directivity characteristics of the array. For multi-layer arrays, a layered calculation and superposition method is used. Based on the different initial positions of each layer, the horizontal beam characteristics of each layer are calculated. The array beam characteristics of the multi-layer array are calculated by energy superposition, as shown below: ; Where H represents the total horizontal beam characteristics of the multilayer array, h n The horizontal beam characteristics of the nth layer of the array are used. The basis for using energy superposition is that the vertical spacing of the three-layer array is much larger than the signal wavelength, and the received signals of each layer are incoherent. If the complex amplitude is directly superimposed, it will introduce unnecessary phase interference, resulting in distortion of the calculation results.

[0022] After determining parameters such as the equipment's operating frequency band, initial array element positions, beamforming angle, and receiving angle, array horizontal beam characteristic analysis can be pre-calculated. The pre-calculated results are then bound to the equipment, and the simulation can be performed by directly calling the pre-calculated beam characteristics, significantly improving the signal-level simulation speed. Two core factors contributing to this speed improvement are: first, the pre-calculated beam characteristics bound to the equipment, used for direct simulation calls, avoid repeated matrix operations in each simulation, reducing real-time computation; second, this beam domain signal simulation method only requires calculating the sound field once per target—the sound field between the array reference element and the target—avoiding the problem of excessively long sound field calculation time caused by a large number of array elements, fundamentally decoupling simulation calculation from the number of array elements.

[0023] The sonar array structure used in this simulation method is a three-layer circular array, such as... Figure 2 As shown, the circular array has a radius of 3m, 512 elements per layer, and a total of 1536 elements. The sonar operates in the 10-3000Hz frequency band. To improve the precision of the array's low-frequency horizontal beam characteristics, 84 frequency points were selected based on 0.1 octave bands across the entire frequency band to calculate the array's horizontal beam characteristics, as shown below. Figure 3 As shown; for the 10Hz to 100Hz frequency band where the line spectrum features of underwater targets are concentrated, the frequency point spacing is adjusted to 0.05 octaves to increase the sampling density. This is because the beamwidth of the low-frequency band is wider, which has a greater impact on the simulation accuracy, and the line spectrum features of the target are mainly concentrated in this frequency band. Increasing the sampling density can more accurately simulate the reception characteristics of the line spectrum signal, such as... Figure 4 As shown.

[0024] The selection of the number of beams falls into two categories. One is to set the number and angle of the beams based on the actual sonar beams, ensuring consistency with the hardware beam configuration of the real equipment and guaranteeing that the simulation results match the output characteristics of the actual equipment. The other is to customize the sonar, setting the number of beams and their corresponding angles according to the user's specific needs, to meet the special requirements of different simulation scenarios. The beamforming angle number N=180, corresponding to a 2° horizontal beam spacing, consistent with the actual hardware configuration of this type of sonar. The recommended setting principle for the number of receiving angles is a 0.1° step, ensuring that the computational load is within an acceptable range while maintaining receiving angle accuracy. The accuracy requirements for azimuth resolution and the improvement or reduction of receiving angle accuracy based on hardware capabilities allow for adjustments. A 0.1° increment can meet the azimuth resolution needs of most underwater targets. For higher accuracy, the increment can be adjusted to 0.05°, and for only rough simulation, it can be adjusted to 0.2°. The number of high-precision receiving angles is K=3600, corresponding to a 0.1° increment. The frequency selection principle within the working frequency band is based on the fact that the target's spectral characteristics are reflected in the low-frequency range. Therefore, more frequency points are selected in the low-frequency band using an octave band method. The more frequency points selected, the longer the pre-calculation time. The number of frequency points can be adjusted according to requirements and hardware capabilities to achieve a balance between accuracy and calculation time.

[0025] Finally, the calculated total horizontal beam characteristic matrix is ​​bound to the unique identifier of the corresponding sonar equipment and stored in the equipment database. The pre-calculation process takes about 16.7 seconds on a regular desktop computer. The result only needs to be calculated once and can be repeatedly called in all simulation scenarios using the equipment without consuming the time resources of real-time simulation.

[0026] II. Beam Domain Signal Simulation Process: Beam domain signal simulation consists of six parts: array beam characteristic pre-calculation, situational scenario design, source generation, channel synthesis, reference signal generation, and beam characteristic superposition. Each part is executed sequentially to collectively generate the beam domain signal. Figure 1 As shown.

[0027] In the array beam characteristic pre-calculation section, the relevant calculation parameters of the equipment are configured, and the array horizontal beam characteristics can be calculated based on the above array horizontal beam characteristic analysis method. After the calculation is completed, it is bound to the equipment and stored in the equipment database. This step is a one-time operation, and all subsequent simulations can directly call it.

[0028] The situational awareness section involves editing the simulation situation based on the established scenario. This includes defining the platform's heading, speed, number of targets, target heading, speed, relative distance, and relative bearing. For different beam signal verification content, corresponding situations are defined to provide situational parameters for beam domain signal simulation. These parameters directly determine the signal's propagation characteristics, such as delay and Doppler shift. Passive single-target simulation, active single-target simulation, and passive multi-target simulation are conducted sequentially. The passive single-target situational awareness is set as follows: the friendly platform's initial position is the origin of the Cartesian coordinate system, heading due north, with a speed of 18 knots; the target platform's initial position is 2000 meters due east, heading due north, with a speed of 18 knots, and it is sailing parallel to the friendly platform; the target platform's physical parameters are a length of 100 meters, a width of 20 meters, and a tonnage of 3000 tons. These parameters directly affect the passive beam domain simulation. Noise generation; Active single-target situation setting: the initial position of the friendly platform is the origin, heading due north, speed 10 knots; the initial position of the target platform is 2100 meters due south, heading due south, speed 10 knots, the two platforms move towards each other, relative speed 20 knots, this relative speed will produce a significant Doppler frequency shift effect; Passive multi-target situation setting: the initial position of the friendly platform is the origin, heading due north, speed 18 knots; a total of 4 targets are set: target 1 is located 3000 meters due east, heading due north, speed 18 knots; target 2 is located 3000 meters due north, heading due west, speed 15 knots; target 3 is located 3000 meters due west, heading due south, speed 12 knots; target 4 is located 3000 meters due south, heading due east, speed 20 knots. The multi-target situation is used to verify the ability of this method to simulate multiple targets simultaneously.

[0029] The signal source generation section includes active transmission signal sources and passive radiated noise sources. Active transmission signal sources include single-frequency pulse signals, linear frequency modulated (LFM) signals, and hyperbolic frequency modulated (HFM) signals, each suitable for different active detection scenarios. Single-frequency pulse signals offer high ranging accuracy, LFM signals provide good range resolution, and HFM signals are suitable for high-speed target detection. The passive radiated noise source generation section uses a general ship radiated noise simulation model. First, based on target physical dimensions, tonnage, speed, and other parameters, the continuous spectrum and line spectrum levels of the platform's radiated noise are simulated. The continuous spectrum mainly consists of mechanical noise and propeller noise, and the spectral levels increase with frequency. The reduction of line spectrum corresponds to the characteristic frequencies of rotating components such as engines and gearboxes, which is an important basis for passive detection. Then, based on the spectral level fitting of the platform radiated noise signal, the typical ship radiated noise simulation is realized, mainly reflecting the continuous spectrum and line spectrum characteristics of the platform radiated noise. The passive signal sampling frequency is 10000Hz, and the signal length of each frame is 10 seconds. The active signal parameters are as follows: single-frequency pulse signal center frequency 2000Hz, pulse width 1 second, pulse period 10 seconds, and emission source level 210dB; linear frequency modulated signal center frequency 2000Hz, bandwidth 500Hz, pulse width 1 second, pulse period 10 seconds, and emission source level 210dB.

[0030] The channel synthesis section calculates sound pressure levels and arrival structures based on the acoustic propagation model, fits the time-domain channel, and reflects the key characteristics of propagation loss and multipath effects in underwater acoustic channels. These two characteristics are the core features that distinguish underwater acoustic channels from other channels. A ray model is used to calculate acoustic propagation characteristics. By decomposing the acoustic signal into multiple independently propagating sound rays, the propagation path, propagation time, propagation loss, angle of arrival, and phase change of each sound ray are obtained by solving the equations, forming a complete sound ray arrival structure. Subsequently, a broadband simulated channel is synthesized using the sound ray arrival structure. The time-domain responses of each sound ray are superimposed to obtain the channel impulse response. The time-domain response is the impulse function, the amplitude is the reciprocal of the propagation loss, and the time delay is the propagation time. Considering the differences in propagation characteristics of different frequency components, broadband channel synthesis is completed across the entire operating frequency band to obtain the frequency-dependent channel transfer function, accurately simulating the frequency-selective fading characteristics of the underwater acoustic channel.

[0031] In the reference signal generation section, the positions of reference array elements are determined based on the initial array position, and the acoustic propagation results between the reference array elements and the target are calculated. The geometric center element of the array is selected as the reference array element. For single-target simulation, only the acoustic field between the reference array element and the target needs to be calculated, eliminating the need to calculate the acoustic field of each element individually. For multi-target simulation, only the acoustic field of the reference array element needs to be calculated once for each target. The computational load is proportional to the number of targets and independent of the number of array elements, which is one of the core advantages of this method compared to traditional element-domain methods. Based on situational information, the slant range and radial velocity of the target relative to the reference array element are calculated in real time. The slant range corresponds to the acoustic signal propagation delay, and the radial velocity corresponds to the Doppler frequency shift. These propagation effects are superimposed on the source signal to obtain the source signal after delay and Doppler correction. Based on the calculation results of the active and passive signal sources and the synthesized underwater acoustic channel from the reference array element to the target, a convolution operation is performed to obtain the time-domain signal of the reference array element, as shown in the following formula: In the formula, For the reference array element time-domain signal, conv is the convolution operation. The source signal is the result of superimposing the time delay effect and the Doppler frequency shift effect. As a reference to the underwater acoustic channel synthesis result from the array element to the target, this operation simulates the propagation process of the acoustic signal in the underwater acoustic channel. The channel multipath effect will cause the signal to generate multiple echoes, and the convolution operation can accurately simulate this process.

[0032] For the beam characteristic superposition part, firstly, a Fourier transform is performed on the time-domain reference signal to obtain the frequency-domain reference signal, and then the frequency-domain reference signal is expanded into a beam frequency-domain reference signal, as shown in the following equation: In the formula, The frequency domain reference signal is the beam. FFT is the Fast Fourier Transform operation. ones(N,1) is an N-row, 1-column matrix of all 1s, where N is the number of beams. The superscript T indicates the matrix transpose operation. This operation realizes the replication of the single-channel frequency domain reference signal to multiple channels. Each beam uses the same reference signal and is subsequently distinguished by different weighting coefficients. This is because the reference signals of all beams are the received signals of the reference array elements, only the array gain for different beams is different.

[0033] Based on the target's azimuth, the array's broadband beam characteristics are... The beam characteristics corresponding to the receiving angle are extracted, and the target azimuth beam characteristic matrix is ​​shown in the following formula: ; in, The target azimuth beam characteristics are defined by N, the number of frequency points for beam characteristic analysis, and K, the number of beams. The length is consistent with the number of frequency points Nf. Each element corresponds to the beam gain of a frequency point. This vector describes the broadband gain characteristics of the array in the target azimuth.

[0034] Target azimuth beam characteristics Interpolation is performed to interpolate the target azimuth beam characteristics to the beam characteristics of the array's operating frequency band in 0.1Hz increments, expressed as follows: Since the pre-calculated beam characteristics are based on 84 discrete frequency points, while the frequency interval of the frequency domain reference signal obtained by the Fast Fourier Transform is 0.1Hz, the sampling frequency is 10000Hz, and the Fast Fourier Transform length is 100000 points, linear interpolation is performed on the target azimuth beam characteristic vector to obtain a continuous beam characteristic vector with a step of 0.1Hz within the array's operating frequency band. Its length is consistent with the frequency domain reference signal, ensuring that each frequency point has a corresponding beam gain value.

[0035] Considering that the channel gain for the signal is reflected through convolution, i.e., frequency domain dot product, and that the array beam characteristics can also be considered as the gains of different beams, in summary, by multiplying the beam frequency domain reference signal with the corresponding frequency array beam characteristics, the beam frequency domain signal can be obtained. This operation transforms the time-domain convolution operation into a frequency-domain dot product operation, significantly improving computational efficiency, as shown in the following equation: ; in, For beam frequency domain signals, , These are the lower and upper operating frequencies of the sonar, respectively. The time-domain signal of the beam can be calculated by performing an inverse Fourier transform on the beam frequency domain signal. This completes the synthesis of beam domain signals.

[0036] III. Verification of the correctness of beam-domain and time-domain signals: To investigate the reliability of this simulation method, passive detection simulation scenarios and active detection simulation scenarios were constructed respectively, and the basic characteristics of the simulation results of passive beam domain received signals and active beam domain echo signals were verified.

[0037] Other simulation parameters are shown in the table below: In the passive detection simulation scenario, the main parameters for platform radiated noise simulation are as follows: target length 100m, target width 20m, target tonnage 3000t, target speed 18kN; other simulation parameters are as follows: number of array elements 1536, lower limit frequency 10Hz, upper limit frequency 3000Hz, sampling frequency 10000Hz, number of beam angles 180, number of receiving angles 3600, number of frames 10, frame length 10s, platform speed 18kN; firstly, the correctness of the energy level simulation results and the fitted source signal is verified by comparing the power spectrum of the passive source fitted signal with the energy level simulation power spectrum, such as... Figure 6 As shown, the two curves have a high degree of overlap, and the slope of the continuous spectrum and the position and amplitude of the line spectrum are consistent, verifying the correctness of the source signal fitting. This verification ensures that the output of the source generation module meets expectations. Next, the correctness of the source signal and beam domain signal is verified. The signal of the beam corresponding to the target azimuth is extracted and subjected to LOFAR spectrum analysis. Comparison with the original source signal line spectrum information shows that the line spectrum position and amplitude in the LOFAR spectrum correspond to the original source signal line spectrum information, with no significant offset or missing values. The correctness of the beam domain signal simulation results is verified. Figure 7 As shown, this verification ensures that no signal distortion is introduced during the beam characteristic superposition process; then, the correctness of the single-target azimuth is verified. The single-target case situation is that the friendly platform is flying straight north at a speed of 18 knots, and the target is located parallel to the friendly platform from the east. The single-target beam domain simulation results are processed, and the single-target azimuth history diagram and frequency azimuth diagram are analyzed, such as... Figure 8 and Figure 9 As shown in the position history diagram, the single-target position history diagram results correspond to the scenario in this case. The target is located at 90° of the platform's hull angle. The frequency position diagram of the last frame of the beam domain signal shows that the positions of each target correspond to the scenario in this case, and the beam domain signal receiving frequency is consistent with the equipment's operating frequency band. This verification ensures the accuracy of the target position information simulation. Finally, the accuracy of the multi-target position verification is performed, such as... Figure 10 As shown, the multi-target case scenario involves the platform flying directly north, with four targets positioned in different azimuths and headings, all at a speed of 18 knots. The multi-target beam domain simulation results are processed, and the multi-target azimuth history map and frequency azimuth map are analyzed, as follows: Figure 11 and Figure 12As shown in the figure, the azimuth history diagram shows that the results of the multi-target azimuth history diagram correspond to the scenario in this case. The frequency azimuth diagram of the last frame of the beam domain signal shows that the target azimuth corresponds to the scenario in this case, with no target confusion or false targets generated. This verification ensures the simulation capability of this method for multi-target scenarios.

[0038] As shown in the table above, the main parameters of the active source transmission signal simulation in the active detection simulation scenario are as follows: the source level of both the single-frequency pulse signal and the linear frequency modulated signal is 210dB, the bandwidth of the linear frequency modulated signal is 500Hz, the center frequency is 2000Hz, the pulse width is 1s, the pulse period is 10s, and the sampling frequency is 10000Hz. First, the correctness of the Doppler frequency shift is verified. When there is relative motion between the platform and the target, the active echo signal exhibits the Doppler effect. The scenario in this case is as follows: Figure 13 The image shows the platform and target moving towards each other at a speed of 10kN. The correctness of the Doppler effect of the beam-domain active echo signal is verified using a single-frequency pulse signal. The time-frequency plot of nine consecutive frames shows that the radial velocity remains constant. Figure 14 As shown, the echo frequencies are consistent across 9 consecutive frames. Based on the Doppler effect calculation formula: ; in, The received frequency after frequency shifting. The original frequency of the actively transmitted signal, c is the underwater sound speed of 1500 m / s, and v o Let v be the velocity of the platform relative to the medium. If the platform is moving towards the target, then v o The operator before is +, and the operator after is -. s Let v be the velocity of the target relative to the medium. If the target is moving towards the platform, then v s The operator is - for the first sign and + for the second; the calculated theoretical receiving frequency is approximately 2027.5Hz, which is basically consistent with the echo frequency of 2026Hz obtained from time-frequency analysis. The error is within the allowable range and conforms to the Doppler frequency shift effect. This verification ensures the correctness of the Doppler effect simulation of the moving target; next, the correctness of the time-frequency characteristics is verified according to the scenario as follows. Figure 13 As shown, the time-frequency characteristics of the beam domain active echo signal are verified using a linear frequency modulated signal; the time-frequency characteristics of the active echo signal corresponding to the target azimuth are analyzed using the short-time Fourier transform method, such as... Figure 15 As shown, the time-frequency diagram indicates a signal frequency range of 1750Hz to 2250Hz and a pulse width of 1s, consistent with the center frequency, bandwidth, and pulse width configured for a linear frequency modulated signal. This verification ensures that the time-frequency characteristics of the active signal are not distorted. Finally, the azimuth and range accuracy is verified according to the hypothetical situation. Figure 13 As shown, the azimuth and range accuracy of the active echo signal in the beam domain is verified using a linear frequency modulated signal. The first frame of the full-beam time-domain signal is extracted for azimuth and range analysis, which involves performing matched filtering on the received signals of each beam. The matched filtering algorithm is shown in the following formula: ; Where fliplr is the flip operation and H is the conjugate operation. For each beam signal, This is the sequence of transmitted signals. The results are shown in the matched filtering results; the azimuth-distance map processing results are displayed as follows: Figure 16 As shown, the target is located at an azimuth of approximately 90° and a distance of approximately 2100m, which is consistent with the scenario in this simulation case. This verifies the correctness of the azimuth and distance map in the active mode of this simulation method. This verification ensures the accuracy of the simulation of the target's azimuth and distance information in active detection.

[0039] IV. Comparison of beam domain signal simulation duration and complexity analysis: Using the controlled variable method, under the condition that the calculation parameters such as marine simulation situation, platform data, equipment data, and environmental data are all the same, signal simulation was carried out using both the traditional array element domain signal simulation method and the beam domain signal simulation method. The marine simulation situation was that the friendly platform and the two targets were sailing directly east at a speed of 10 knots. The pre-calculation time of the array beam characteristics of this method was 16.7447s, and the average simulation time of a single frame from the first to the tenth frame was about 6.2s. The traditional array element domain signal simulation method had no pre-calculation process, and the average simulation time of a single frame from the first to the third frame was about 456s. It can be observed that this method has a significant advantage in simulation calculation time compared with the traditional array element domain signal simulation method, which is a major breakthrough and greatly solves the problem of excessive calculation time of the traditional array element domain signal simulation method.

[0040] The analysis of computation time and complexity is presented from four aspects: First, in terms of algorithms, traditional array element domain signal simulation methods require sound field calculation, channel synthesis, channel convolution, and Doppler interpolation calculation for each array element; while the time consumption of beam domain signal simulation methods mainly includes array beam characteristic calculation and beam domain signal generation. Beam domain signal generation only requires sound field calculation, channel synthesis, and Doppler interpolation for the reference array element, and then the beam domain signal simulation result is obtained by coupling the reference array element signal with the array horizontal beam characteristics. The increase in the number of array elements has a greater impact on the computational load of traditional array element domain simulation methods, but a smaller impact on the computational load of beam domain simulation methods. Secondly, in terms of pre-calculation, the beam characteristics of the sonar horizontal array can be pre-calculated, and the pre-calculation results can be bound to the equipment and stored in the equipment database. During the signal simulation process, it only needs to be called up, which reduces the signal simulation time to a certain extent. Thirdly, in terms of underlying computation, the traditional array element domain signal simulation method in the control group involves a lot of convolution operations, and the computational complexity is linearly positively correlated with the number of array elements. This beam domain signal simulation method uses frequency domain dot multiplication operations, and the computational complexity is only related to the number of beams and the number of sampling points. When the number of array elements is large, the beam domain signal simulation method has lower computational complexity than the traditional array element domain signal simulation method. Fourthly, regarding subsequent impacts, the beam domain simulation method has already performed beamforming operations on the signal, which can reduce the time required for subsequent signal processing algorithms to some extent. The array element domain signal generated by the traditional method still needs to undergo additional beamforming processing, further increasing the computation time of the entire process.

[0041] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0042] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics, characterized in that, The method includes the following steps: S1. Configure the sonar equipment's operating frequency band, initial position of array elements, number of beamforming angles, and number of high-precision receiving angles. Calculate the array's horizontal beam characteristics and bind the array's horizontal beam characteristics to the corresponding sonar equipment and store them in the equipment database. S2. Construct a marine simulation scenario and obtain simulation equipment platform parameters, gridded marine environment parameters of the target sea area, and marine simulation situation information including platform heading, platform speed, number of targets, target heading, target speed, relative distance, and relative bearing; S3. Generate active transmission signals based on the active operating mode of the sonar, and generate passive platform radiated noise source signals based on the passive operating mode of the sonar. S4. Based on the gridded marine environmental parameters of the target sea area, the sound pressure result and the sound ray arrival structure are calculated using the sound propagation model. The sound ray arrival structure is then used to synthesize a broadband simulation channel that reflects the propagation loss and multipath effect of the underwater acoustic channel. S5. Determine the position of the reference array element. For a single target, only the sound field between the reference array element and the target is calculated. Based on the marine simulation situation information and equipment platform parameters, add delay effect and Doppler frequency shift effect to the active transmission signal or passive platform radiated noise source signal. Perform convolution operation on the source signal after superposition effect with the broadband simulation channel to obtain the time domain reference signal of the reference array element. S6. Based on the time-domain reference signal of the reference array element and the pre-calculated and stored array horizontal beam characteristics, the beam-domain time-domain signal is obtained through frequency-domain directivity weighting.

2. The beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics according to claim 1, characterized in that, The pre-calculation of the array horizontal beam characteristics in step S1 specifically includes: S11. Calculate the array manifold matrix based on the sonar equipment parameters, where the column vectors of the array manifold matrix are the response vectors of the array to different beamforming angles; S12. Calculate the product of the array manifold matrix corresponding to the number of beamforming angles and the array manifold matrix corresponding to the number of high-precision receiving angles to obtain the beam characteristic matrix of the narrowband signal. S13. For broadband signals, select multiple frequency points according to a preset octave interval, calculate the beam characteristic matrix corresponding to each frequency point, and splice them together to obtain the broadband array beam characteristic matrix. S14. For multi-layer arrays, calculate the horizontal beam characteristics of each layer based on its initial position, and obtain the total horizontal beam characteristics of the multi-layer array by energy superposition.

3. The beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics according to claim 2, characterized in that, In step S11, for a uniformly distributed linear array containing M elements, the angle formed by the array with respect to the k-th beam is... The response vector is: In the formula, d is the element spacing, λ is the wavelength of the acoustic signal, and the superscript T indicates the matrix transpose operation.

4. The beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics according to claim 2, characterized in that, In step S13, frequency points are selected at intervals of 0.1 octaves of the lower limit frequency, and the frequency point selection density is increased for the low-frequency bands in the target line spectrum feature set; the total number of frequency points is dynamically adjusted according to the simulation accuracy requirements and the computing power of the computer hardware.

5. The beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics according to claim 1, characterized in that, In step S1, the number of beamforming angles is set according to the actual number of beams and corresponding angles configured in the sonar equipment, or customized according to the user's simulation requirements; the number of high-precision receiving angles is set in 0.1° increments, or the increment precision is adjusted according to the azimuth resolution requirements and hardware level.

6. The beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics according to claim 1, characterized in that, In step S3, the actively transmitted signal can be any one of a single-frequency pulse signal, a linear frequency modulated signal, or a hyperbolic frequency modulated signal; the passive platform radiated noise source is generated using a general ship radiated noise simulation model, specifically: Based on the target's physical dimensions, tonnage, and speed parameters, the continuous spectrum and line spectrum levels of the platform's radiated noise are simulated and calculated. The platform's radiated noise time-domain signal is obtained by fitting the continuous spectrum and line spectrum levels, where the platform's radiated noise time-domain signal includes both continuous spectrum and line spectrum features.

7. The beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics according to claim 1, characterized in that, In step S4, the sound propagation model adopts the ray model, and the broadband simulation channel is synthesized by the time-domain superposition of the sound rays arriving at the structure.

8. The beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics according to claim 1, characterized in that, In step S5, the formula for calculating the time-domain reference signal of the reference array element is: In the formula, For the reference array element time-domain signal, conv is the convolution operation. The source signal is the result of superimposing the time delay effect and the Doppler frequency shift effect. The result is the underwater acoustic channel synthesis from the reference array element to the target.

9. The beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics according to claim 1, characterized in that, Step S6, beam domain signal synthesis, specifically includes: S61. Perform a fast Fourier transform on the time-domain reference signal of the reference array element to obtain the frequency-domain reference signal, and then expand the frequency-domain reference signal into a dimension-matched beam frequency-domain reference signal. S62. Based on the actual azimuth of the detected target, extract the target azimuth beam characteristic matrix corresponding to the receiving angle from the pre-calculated and stored broadband array beam characteristic matrix. S63. Perform linear interpolation on the target azimuth beam characteristic matrix to obtain a continuous beam characteristic matrix with a step of 0.1Hz within the array's operating frequency band. S64. Multiply the beam frequency domain reference signal with the interpolated continuous beam characteristic matrix point by point in the frequency domain to obtain the beam frequency domain signal. S65. Perform a fast inverse Fourier transform on the beam frequency domain signal to obtain the final beam time domain signal.

10. The beam domain signal simulation method based on pre-calculation of array horizontal beam characteristics according to claim 9, characterized in that, In step S61, the formula for calculating the beam frequency domain reference signal is: In the formula, The frequency domain reference signal is the beam, FFT is the fast Fourier transform operation, ones(N,1) is an N-row, 1-column matrix of all 1s, N is the number of beams, and the superscript T indicates the matrix transpose operation.

Images