A ship active signal time-frequency analysis method based on time-frequency parameter fine estimation

Abstract

The application discloses a ship active signal time-frequency analysis method based on time-frequency parameter fine estimation, and comprises the following steps: (1) obtaining a ship active signal to be processed; (2) calculating a single-frequency signal energy fluctuation characteristic of the ship active signal to be processed; (3) calculating a received single-frequency signal time existing range based on the single-frequency signal energy fluctuation characteristic; (4) calculating a received single-frequency signal frequency based on the received single-frequency signal time existing range; and (5) calculating a received frequency modulation signal time-frequency distribution based on the received single-frequency signal frequency. The method is suitable for a low signal-to-noise ratio condition, can effectively calculate a time-frequency distribution of a frequency modulation signal in a two-component ship active signal composed of the frequency modulation signal and a single-frequency signal, and provides a basis for target feature extraction based on the ship active signal.

Description

Technical Field

[0001] This invention belongs to the field of ship signal processing, specifically relating to a time-frequency analysis method for ship active signals based on fine estimation of time-frequency parameters. Background Technology

[0002] Active ship signals include single-frequency, linear frequency modulation (FM), hyperbolic frequency modulation (HFM), and combined waveform active ship signals composed of the above three single-waveform signals. To achieve effective target detection and feature extraction, a common combination of active ship signals is a two-component form consisting of a single FM or hyperbolic frequency modulation signal and a single-frequency signal connected end-to-end. The FM signal is primarily used for target detection, while the single-frequency signal is used for feature extraction. Specifically, because the FM signal is non-stationary, time-frequency analysis can extract fine target features.

[0003] Time-frequency analysis methods are divided into two main categories: (1) Bilinear time-frequency analysis methods, mainly including Wigner-Willi distribution, etc. Although this type of method has high resolution, it has a large computational load, and there are cross terms when analyzing hyperbolic frequency modulated signals, which affects the effective extraction of features. (2) Linear time-frequency analysis methods, mainly including short-time Fourier transform, etc. This type of method has low computational complexity, but time and frequency resolution are limited, which also affects the extraction performance of target features. Summary of the Invention

[0004] Technical problem: The purpose of this invention is to provide a time-frequency analysis method for active signals from ships. This method can effectively estimate the time-frequency parameters of a received single-frequency signal when the parameters of the transmitted signal are known, and design the time-frequency distribution parameters of the frequency-modulated signal based on this, so as to obtain a time-frequency distribution of the frequency-modulated signal with better time and frequency resolution, thus providing a basis for the effective extraction of target features.

[0005] Technical solutions, such as... Figure 1 As shown, this invention proposes a time-frequency analysis method for ship active signals based on fine estimation of time-frequency parameters. The method includes the following steps:

[0006] (1) Obtain the active signal x(n) of the ship to be processed;

[0007] (2) Calculate the single-frequency signal energy fluctuation characteristics E of the active signal x(n) of the ship to be processed. CW (m CW );

[0008] (3) Based on the energy fluctuation characteristics E of single-frequency signals CW (m CW ), calculate the time range n of the received single-frequency signal. CW,recv ;

[0009] (4) Based on the time range n of the received single-frequency signal CW,recv Calculate the frequency f of the received single-frequency signal. CW,recv ;

[0010] (5) Based on the received single-frequency signal frequency f CW,recv Calculate the time-frequency distribution TF of the received FM signal. FM (m FM ,l FM ).

[0011] Furthermore, in step (1), the active signal x(n) of the ship to be processed is obtained using the following method, specifically including the following steps:

[0012] The ship's active signal x(n) to be processed is obtained from real-time data collected from N sampling points of the ship's underwater acoustic sensor, where n = 0, 1, …, N-1; or data from N sampling points is extracted from memory as the ship's active signal x(n), where n = 0, 1, …, N-1, and n is the time index of the ship's active signal x(n) to be processed, and N is the number of signal sampling points, which is N=2. κ κ is a positive integer greater than or equal to 10.

[0013] Furthermore, in step (2), the energy fluctuation characteristics E of the single-frequency signal are calculated using the following method. CW (m CW Specifically, it includes the following steps:

[0014] (2-1) Initialize the parameters for calculating the energy fluctuation characteristics of a single-frequency signal, specifically including the initialization of the following parameters:

[0015] ① Transmit single-frequency signal frequency f CW,trans Initialized to: 0 <f CW,trans <f s / 2 positive real numbers, where f s It is the sampling frequency;

[0016] ② Transmit single-frequency signal pulse width N CW,trans Initialized to: 0 <N CW,trans Positive integers ≤ N / 2;

[0017] ③ The single-frequency signal detection window length scaling parameter α is initialized to a positive real number where 0 < α ≤ 1;

[0018] ④ Single-frequency signal detection window length N CW,window Initialized to: N CW,window =max{4, round{αN CW,trans A positive integer, where max{·} represents the maximum value function and round{·} represents the rounding function;

[0019] ⑤ single-frequency signal detection step N CW,step initialized to: 1 ≤ N CW,step ≤ floor{N CW,window / 4}, where floor{·} denotes the floor function;

[0020] ⑥ total number of frames M in single-frequency signal detection CW initialized to: M CW = floor{(N-N CW,window ) / N CW,step}-1, where floor{·} denotes the floor function;

[0021] ⑦ target maximum speed v max initialized to: 10 ≤ v max ≤ 20, where v CW,det is a positive real number;

[0022] ⑧ underwater average sound speed c0initialized to: 1400 ≤ c0≤ 1600, where c0is a positive real number;

[0023] (2-2) Calculate the single-frequency signal detection frequency range f CW,det :

[0024]

[0025] where η min and η max are the minimum and maximum Doppler factors corresponding to the target maximum speed v max , i.e., η min = (c0-v max ) / (c0+v max ) and η max = (c0+v max ) / (c0-v max );

[0026] (2-3) Calculate the single-frequency signal detection frequency point starting index L CW,start and the total number of single-frequency signal detection frequency points L CW,total :

[0027]

[0028]

[0029] where f CW,det (1) and f CW,det (2) represent the 1st and 2nd elements of the single-frequency signal detection frequency range f CW,det , respectively;

[0030] (2-4) Extract the single-frequency signal detection segmented data sequence y CW,0 (rCW ),y CW,1 (r CW ),…, ,…, , where the mth CW Segmented data sequence for single-frequency signal detection for:

[0031]

[0032] in, For single-frequency signal detection segmented time index, For single-frequency signal detection, the frame index is... =0,1,…,M CW -1;

[0033] (2-5) Calculate the time-frequency distribution TF of the single-frequency signal detection CW (m CW ,l CW ):

[0034]

[0035] Among them, l CW This is the frequency index for single-frequency signal detection, where j is the imaginary unit, i.e. , |·| is the modulo function;

[0036] (2-6) Calculate the energy fluctuation characteristics E of a single-frequency signal CW (m CW ):

[0037] .

[0038] Furthermore, in step (3), the time range n of the received single-frequency signal is calculated using the following method. CW,recv Specifically, it includes the following steps:

[0039] (3-1) The relative amplitude parameter β for single-frequency signal bandwidth search is initialized to a positive real number where 0 < β < 1;

[0040] (3-2) Calculate the energy fluctuation characteristics E of a single-frequency signal CW (m CW The time frame index m0 corresponding to the maximum value:

[0041]

[0042] in, E represents the energy fluctuation characteristics of a single-frequency signal. CW (m CW The time frame index corresponding to the maximum value;

[0043] (3-3) Calculate the bandwidth amplitude E of a single-frequency signal B :

[0044] ;

[0045] (3-4) Search for the energy fluctuation characteristics of a single-frequency signal E CW (m CW The maximum value to the left is less than the bandwidth amplitude E of a single-frequency signal. B Time frame index set m left :

[0046]

[0047] in, It represents all that satisfy E CW (m CW )≤E B Single-frequency signal detection time frame index m CW The set consisting of max{·} represents the function that takes the maximum value;

[0048] (3-5) Calculate the frame start index m of a single-frequency signal CW,left :

[0049]

[0050] in, Denotes the empty set, m left (end) represents the time frame index set m left The last element;

[0051] (3-6) Search for the energy fluctuation characteristics of a single-frequency signal E CW (m CW The maximum value to the right is less than the bandwidth amplitude E of a single-frequency signal. B Time frame index set m right :

[0052]

[0053] Where min{·} represents the minimum value function;

[0054] (3-7) Calculate the frame cutoff index m of a single-frequency signal CW,right :

[0055]

[0056] Where, m right (1) Represents the time frame index set m right The first element;

[0057] (3-8) Calculate the time range n of the received single-frequency signalCW,recv :

[0058]

[0059] in, Indicates search The time index n corresponding to the minimum value, Indicates search The time index n corresponding to the minimum value.

[0060] Furthermore, in step (4), the frequency f of the received single-frequency signal is calculated using the following method. CW,recv Specifically, it includes the following steps:

[0061] (4-1) Initialize the parameters for calculating the frequency of the received single-frequency signal, specifically including the initialization of the following parameters:

[0062] ① Transmit single-frequency signal period N CW,period Initialized to: N CW,period =round{f s / f CW,trans A positive integer f, where f s This is the sampling frequency, and round{·} represents the rounding function.

[0063] ② Traverse the set of frequency detectors for pulse width single-frequency signals f CW,trav Initialize to: Its length is N CW,period A vector of all zeros;

[0064] ③ Traverse the set of peak values ​​of the amplitude spectrum of the pulse width single-frequency signal A CW,trav Initialize to: Its length is N CW,period A vector of all zeros;

[0065] ④ Traverse the pulse width index d trav Initialized to: d trav =0;

[0066] (4-2) Calculate the d-th ... trav A single-frequency traversal pulse width signal :

[0067]

[0068] in, It is the dth trav A time index for traversing a single-frequency pulse width signal. It is the dth trav The number of sampling points for a single-frequency traversal pulse width signal, n CW,recv (1) and n CW,recv(2) respectively represent the time range n of receiving a single-frequency signal. CW,recv The first and second elements;

[0069] (4-3) Calculate the d-th ... trav A single-frequency traversal pulse width signal amplitude spectrum :

[0070]

[0071] in, It is the dth trav A frequency index for a single-frequency pulse width signal is traversed, where j is the imaginary unit, i.e. , |·| is the modulo function;

[0072] (4-4) Update the set of frequency detectors f for traversing pulse width single-frequency signals CW,trav The dth trav element f CW,trav (d trav ):

[0073]

[0074] in, Indicates the search amplitude spectrum Frequency index corresponding to the maximum value ;

[0075] (4-5) Update the set of peak values ​​of the amplitude spectrum of the pulse width single-frequency signal A. CW,trav The dth trav Element A CW,trav (d trav ):

[0076]

[0077] Where max{·} represents the function that takes the maximum value;

[0078] (4-6) Update the traversal pulse width index d trav :

[0079] ;

[0080] (4-7) Determine the pulse width index d during traversal trav Does the traversal deadline condition apply?

[0081]

[0082] If the condition is met, proceed to step (4-8); otherwise, return to step (4-2).

[0083] (4-8) Calculate the frequency f of the received single-frequency signal. CW,recv :

[0084]

[0085] Where d0 is the frequency f of the received single-frequency signal. CW,recv The corresponding traversal pulse width peak index, A CW,trav (d trav ) represents the set of peak values ​​of the amplitude spectrum of a pulse-width single-frequency signal. CW,trav The dth trav elements, f CW,trav (d0) represents the set of frequency detectors for traversing pulse width single-frequency signals f. CW,trav The d0th element.

[0086] Furthermore, in step (5), the time-frequency distribution TF of the received FM signal is calculated using the following method. FM (m FM ,l FM Specifically, it includes the following steps:

[0087] (5-1) Initialize the parameters for calculating the time-frequency distribution of the received FM signal, specifically including the initialization of the following parameters:

[0088] ① Transmit the starting frequency f of the FM signal FM,1,trans Initialized to: 0 <f FM,1,trans <f s / 2 positive real numbers, where f s It is the sampling frequency;

[0089] ② The cutoff frequency f of the transmitted FM signal FM,2,trans Initialized to: 0 <f FM,2,trans <f s Positive real numbers of 2 / 2;

[0090] ③ Transmit FM signal pulse width N FM,trans Initialized to: 0 <N FM,trans ≤NN CW,trans Positive real numbers;

[0091] ④ Time-frequency distribution frequency tolerance parameter λ f Initialize to: 0 < λ f Positive real numbers ≤ 1;

[0092] ⑤ Time-frequency distribution time tolerance parameter λ t Initialize to: 0 < λ t Positive real numbers ≤ 1;

[0093] (5-2) Calculate the frequency range f of the received FM signal. FM,recv :

[0094]

[0095] where η0= f CW,recv CW,trans is the target detection Doppler factor, min{·} and max{·} represent the minimum and maximum functions, respectively;

[0096] (5-3) Calculate the time existing range n FM,recv :

[0097]

[0098] where n CW,recv (1) and n CW,recv (2) represent the 1st and 2nd elements of the time existing range n CW,recv of the received single-frequency signal, respectively;

[0099] (5-4) Calculate the window length N FM,window and the step N FM,step of the time-frequency distribution of the received frequency-modulated signal:

[0100]

[0101] where f FM,recv (1) and f FM,recv (2) represent the 1st and 2nd elements of the frequency range f FM,recv of the received frequency-modulated signal, respectively, round{·} represents the rounding function, and floor{·} represents the floor function;

[0102] (5-5) Calculate the starting index L FM,start and the total number L FM,total of frequency points of the received frequency-modulated signal:

[0103]

[0104] where f FM,range is the frequency width range of the received frequency-modulated signal, the 1st element f FM,range (1) of f FM,range = min{f s / 2, max{0, (1+λ f ) f FM,recv (1)- λ f f FM,recv (2)} and the 2nd element f FM,range (2) of f FM,range = min{f s / 2, (1+λ f ) f FM,recv (2)- λ​f f FM,recv (1)},f FM,recv (1) and f FM,recv (2) represent the frequency range f of the received FM signal respectively. FM,recv The first and second elements;

[0105] (5-6) Calculate the total number of frames M when the received FM signal time-frequency distribution is calculated. FM :

[0106]

[0107] (5-7) Extract the segmented data sequence y of the received FM signal FM,0 (r FM ), y FM,1 (r FM ), …, ,…, , where the mth FM Segmented data sequence of received FM signal for:

[0108]

[0109] Where, r FM For the time-frequency distribution segmented time index of the received FM signal, m FM For the time-frequency distribution time frame index of the received FM signal, m FM =0,1,…,M FM -1, n FM,recv (1) is the range n of the received FM signal time calculated in step (5-3). FM,recv The first element;

[0110] (5-8) Calculate and output the time-frequency distribution TF of the received FM signal. FM (m FM , l FM ):

[0111]

[0112] Among them, l FM The frequency point index for the time-frequency distribution of the received FM signal, where j is the imaginary unit, i.e. , where |·| is the modulo function.

[0113] Beneficial effects: Compared with the prior art, the technical solution of the present invention has the following beneficial technical effects:

[0114] (1) The present invention makes full use of the frequency and pulse width of the transmitted single-frequency signal and designs the energy fluctuation characteristics of the single-frequency signal, as shown in step (2). This feature contains the transmitted single-frequency signal information and can effectively reduce the impact of noise and other interference on the subsequent calculation of the time range of the received single-frequency signal.

[0115] (2) The present invention makes full use of the period of the transmitted single-frequency signal to realize the calculation of the frequency of the received single-frequency signal, as shown in step (4). This calculation process makes full use of the principle of spectrum leakage of single-frequency signal to calculate the frequency of the received single-frequency signal, which provides a basis for the design of the time-frequency distribution parameters of the subsequent frequency modulation signal.

[0116] (3) The present invention makes full use of the received single-frequency signal frequency, calculates the frequency range of the frequency modulation signal, and further designs the frequency modulation signal time-frequency distribution window length based on this, as shown in step (5). The designed frequency modulation signal time-frequency distribution window length takes into account the trade-off between time and frequency resolution. Using the obtained time-frequency distribution, parameters such as instantaneous frequency can be effectively extracted, which is beneficial for target detection and feature extraction based on ship active signals. Attached Figure Description

[0117] Figure 1 This is a flowchart illustrating the method of the present invention.

[0118] Figure 2 This is a single-frequency signal energy fluctuation characteristic diagram of Example 1.

[0119] Figure 3 This is a time-frequency distribution diagram of the received FM signal in Example 1.

[0120] Figure 4 This is a single-frequency signal energy fluctuation characteristic diagram of Example 2.

[0121] Figure 5 This is a time-frequency distribution diagram of the received FM signal in Example 2. Detailed Implementation

[0122] To better understand the purpose, structure, and function of this invention, the invention will be further described below with reference to the accompanying drawings.

[0123] In an embodiment of the present invention, the simulated received signal model x(t) is set as:

[0124]

[0125] Among them, A FM and A CW The signal amplitudes τ and τ of the frequency-modulated and single-frequency signals are respectively. 0,FM and τ 0,CW τ represents the start time of the frequency-modulated (FM) and single-frequency (SFM) signals, respectively. FM and τCW Let be the pulse widths of the frequency-modulated (FM) and single-frequency (SFM) signals, respectively, and satisfy τ. 0,FM +τ FM =τ 0,CW , τ total ω(t) is the pulse width for signal sampling; ω(t) is the signal with a mean of 0 and a variance of σ. 2 Gaussian white noise, variance σ 2 The magnitude depends on the signal-to-noise ratio (SNR): SNR = 10log 10 [(A FM +A CW ) 2 / (4σ 2 )];φ FM (t) and φ CW (t) represents the instantaneous phase of the frequency-modulated and single-frequency signals, respectively, and is expressed as:

[0126]

[0127]

[0128] Where, φ 0,FM and φ 0,CW The initial phases of the frequency-modulated and single-frequency signals are f and f, respectively. 1,FM and f 1,CW These are the starting frequencies of the frequency-modulated (FM) and single-frequency (SFM) signals, respectively; μ is the modulation frequency of the linear frequency-modulated (LFM) signal, defined as: μ = (f 2,FM -f 1,FM ) / τ FM f 2,FM k is the cutoff frequency of the FM signal; k0 is the slope of the hyperbolic FM signal period, defined as: k0=(f 2,FM -f 1,FM ) / (f 1,FM f 2,FM τ FM ).

[0129] With sampling frequency f s Discrete sampling is performed on the received signal x(t) from the above simulation to obtain the signal sampling data sequence x(n):

[0130]

[0131] Wherein, the number of signal sampling points N = round{τ total f s}; Frequency modulation signal start point N 0,FM = round{τ 0,FM f s The transmitted frequency modulation signal pulse width N FM,trans = round{τ FM fs}, the single frequency signal starting point N 0,CW = round{τ 0,CW f s}, the transmitted single frequency signal pulse width N CW,trans = round{τ CW f s} ; the discrete form of the instantaneous phase of the frequency modulation signal and the single frequency signal φ FM (n / f s ) and φ CW (n / f s ) are expressed as:

[0132]

[0133] .

[0134] Embodiment 1:

[0135] The parameter settings of the simulation signal are as follows: the frequency modulation signal is a hyperbolic frequency modulation signal, the signal amplitude A FM = 1 of the frequency modulation signal, the signal amplitude A CW = 1 of the single frequency signal; the starting time τ 0,FM = 0.5s of the frequency modulation signal, the starting time τ 0,CW = 1.5s of the single frequency signal; the signal pulse width τ FM = 1s of the frequency modulation signal, the signal pulse width τ CW = 0.5s of the single frequency signal, the signal sampling pulse width τ total = 3.2768s; the signal-to-noise ratio SNR = -5dB; the initial phase φ 0,FM = 0 of the frequency modulation signal, the initial phase φ 0,CW = 0 of the single frequency signal; the starting frequency f 1,FM = 2960Hz of the frequency modulation signal, the cutoff frequency f 2,FM = 3450Hz of the frequency modulation signal, the starting frequency f 1,CW = 3750Hz of the single frequency signal; the sampling frequency f s = 10000Hz.

[0136] The time-frequency distribution of the received frequency modulation signal is calculated as follows:

[0137] According to step (1), the data of 32768 sampling points of the signal sampling points are extracted from the memory as the to-be-processed ship active signal x(n);

[0138] According to step (2), the transmitted single frequency signal frequency f CW,trans = 3800 is set, and the transmitted single frequency signal pulse width N CW,trans= 5000, single frequency signal detection window length ratio parameter a = 0.1, single frequency signal detection window length N CW,window = 500, single frequency signal detection step N CW,step = 125, total number of single frequency signal detection frames M CW = 257, target maximum speed v max = 20, average underwater sound speed c0= 1500; calculate single frequency signal detection frequency range f CW,det = [3700, 3900]; calculate single frequency signal detection frequency point starting index L CW,start = 185, total number of single frequency signal detection frequency points L CW,total = 11; extract single frequency signal detection segmented data sequence y CW,0 (r CW ), y CW,1 (r CW ), …, , …, y CW,256 (r CW ); calculate single frequency signal detection time-frequency distribution TF CW (m CW , l CW ); calculate single frequency signal energy fluctuation feature E CW (m CW ), as shown in Figure 2 .

[0139] According to step (3), set the single frequency signal bandwidth search relative amplitude parameter β = 0.707; calculate the index m0 = 126 corresponding to the maximum value of the single frequency signal energy fluctuation feature E CW (m CW ); calculate the single frequency signal bandwidth amplitude E B = 1.1515; calculate the single frequency signal time frame starting index m CW,left = 119; calculate the single frequency signal time frame cutoff index m CW,right = 158; calculate the received single frequency signal time existence range n CW,recv = [15125, 20000].

[0140] According to step (4), set the transmitted single frequency signal period N CW,period = 3, traverse the pulse width single frequency signal frequency detection set , traverse the pulse width single frequency signal amplitude spectrum peak value set , traverse the pulse width index d trav = 0; calculate the d trav th traverse pulse width single frequency signal ; calculate the amplitude spectrum of the d trav th traverse pulse width single frequency signal ​Update the set of frequency detectors f for traversing pulse width single-frequency signals. CW,trav The dth trav element f CW,trav (d trav Update the set of peak values ​​A of the amplitude spectrum of the pulse width single-frequency signal. CW,trav The dth trav Element A CW,trav (d trav Update the traversal pulse width index d. trav ; Calculate the received single-frequency signal frequency f CW,recv =3750.

[0141] According to step (5), set the starting frequency f of the transmitted frequency modulation signal. FM,1,trans =3000, cutoff frequency for transmitted FM signal f FM,2,trans =3500, Transmit FM signal pulse width N FM,trans =10000, time-frequency distribution frequency tolerance parameter λ f =0.15, time-frequency distribution time tolerance parameter λ t =0.15; Calculate the frequency range f of the received FM signal. FM,recv =[2960, 3450]; Calculate the time range n for receiving the FM signal. FM,recv =[5125, 15125]; Calculate the time-frequency distribution window length N of the received FM signal. FM,window =637, calculate the time-frequency distribution step N of the received FM signal. FM,step =159; Calculate the starting index L of the received FM signal frequency point. FM,start =184, calculate the total number of frequency points L for receiving FM signals. FM,total =42; Calculate the total number of time frames M for the received FM signal time-frequency distribution. FM =76; Extract the segmented data sequence y of the received FM signal. FM,0 (r FM ), y FM,1 (r FM ), …, , …, y FM,75 (r FM ); Calculate and output the time-frequency distribution TF of the received FM signal. FM (m FM , l FM ),like Figure 3 As shown.

[0142] Example 2:

[0143] The simulation signal parameters are set as follows: the frequency modulation signal is a linear frequency modulation signal, and the signal amplitude A of the frequency modulation signal is... FM =1, the signal amplitude A of the single-frequency signal CW =1; the start time τ of the frequency modulation signal0,FM =1 s, initial time of the single-frequency signal τ 0,CW =3 s; signal pulse width of the frequency-modulated signal τ FM =2 s, signal pulse width of the single-frequency signal τ CW =1 s, signal sampling pulse width τ total =5.12 s; signal-to-noise ratio SNR=-3 dB; initial phase of the frequency-modulated signal φ 0,FM =0, initial phase of the single-frequency signal φ 0,CW =0; initial frequency of the frequency-modulated signal f 1,FM =225 Hz, cut-off frequency of the frequency-modulated signal f 2,FM =421 Hz, initial frequency of the single-frequency signal f 1,CW =490 Hz; sampling frequency f s =2000 Hz.

[0144] The simulation signal is calculated as follows for the time-frequency distribution of the received frequency-modulated signal:

[0145] According to step (1), the data of 10240 sampling points of the signal are extracted from the memory as the ship active signal x(n) to be processed;

[0146] According to step (2), the frequency f CW,trans =500 of the transmitted single-frequency signal is set; CW,trans =2000, the single-frequency signal detection window length proportion parameter α=0.1, the single-frequency signal detection window length N CW,window =200, the single-frequency signal detection step N CW,step =50, the total number of single-frequency signal detection frames M CW =199, the target maximum speed v max =20, the average underwater sound speed c0=1500; the single-frequency signal detection frequency range f CW,det =[487, 514]; the single-frequency signal detection frequency point starting index L CW,start =49, the total number of single-frequency signal detection frequency points L CW,total =3; the single-frequency signal detection segmented data sequence y CW,0 (r CW ), y CW,1 (r CW ), …, , …, y CW,198 (r CW ); the single-frequency signal detection time-frequency distribution TF CW (m CW , l CW ); the single-frequency signal energy fluctuation feature E CW (m CW ) is calculated, for exampleFigure 4 As shown.

[0147] Based on step (3), set the relative amplitude parameter β=0.707 for the single-frequency signal bandwidth search; calculate the energy fluctuation characteristics E of the single-frequency signal. CW (m CW The index m0 corresponding to the maximum value is 127; calculate the bandwidth amplitude E of the single-frequency signal. B =0.5384; Calculate the frame start index m of a single-frequency signal. CW,left =121; Calculate the frame cutoff index m of the single-frequency signal. CW,right =159; Calculate the time range n for receiving a single-frequency signal. CW,recv =[6150, 8050].

[0148] According to step (4), set the transmission single-frequency signal period N. CW,period =4, Traverse the pulse width single-frequency signal frequency detection set Traverse the set of peak values ​​of the amplitude spectrum of a single-frequency pulse width signal. Traverse pulse width index d trav =0; calculate the dth th trav A single-frequency traversal pulse width signal ; Calculate the dth trav A single-frequency traversal pulse width signal amplitude spectrum Update the set of frequency detectors f for traversing pulse width single-frequency signals. CW,trav The dth trav element f CW,trav (d trav Update the set of peak values ​​A of the amplitude spectrum of the pulse width single-frequency signal. CW,trav The dth trav Element A CW,trav (d trav Update the traversal pulse width index d. trav ; Calculate the received single-frequency signal frequency f CW,recv =490.

[0149] According to step (5), set the starting frequency f of the transmitted frequency modulation signal. FM,1,trans =230, cutoff frequency for transmitted FM signal f FM,2,trans =430, Transmitted FM signal pulse width N FM,trans =4000, time-frequency distribution frequency tolerance parameter λ f =0.15, time-frequency distribution time tolerance parameter λ t =0.15; Calculate the frequency range f of the received FM signal. FM,recv =[225,421];Calculate the time range n of the received FM signal. FM,recv =[2150, 6150]; Calculate the time-frequency distribution window length N of the received FM signal.FM,window =286, calculating the frequency step N of the time-frequency distribution of the received frequency modulation signal FM,step =71; calculating the starting index L of the frequency points of the received frequency modulation signal FM,start =28, calculating the total number L of frequency points of the received frequency modulation signal FM,total =37; calculating the total number M of time frames of the time-frequency distribution of the received frequency modulation signal FM =68; extracting the segmented data sequence y of the received frequency modulation signal FM,0 (r FM ), y FM,1 (r FM ), …, , …, y FM,67 (r FM ); calculating and outputting the time-frequency distribution TF of the received frequency modulation signal FM (m FM , l FM ), as shown in Figure 5 .

[0150] It can be understood that the present application is described through some embodiments, and those skilled in the art know that various changes or equivalent replacements can be made to these features and embodiments without departing from the spirit and scope of the present application. In addition, under the guidance of the present application, modifications can be made to these features and embodiments to adapt to specific conditions and materials without departing from the spirit and scope of the present application. Therefore, the present application is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of the present application are within the scope of protection of the present application.

Claims

1. A time-frequency analysis method for ship active signals based on fine estimation of time-frequency parameters, characterized in that, The method includes the following steps: (1) Obtain the active signal x(n) of the ship to be processed; (2) Calculate the single-frequency signal energy fluctuation characteristics E of the active signal x(n) of the ship to be processed. CW (m CW ); (3) Based on the energy fluctuation characteristics E of single-frequency signals CW (m CW ), calculate the time range n of the received single-frequency signal. CW,recv ; (4) Based on the time range n of the received single-frequency signal CW,recv Calculate the frequency f of the received single-frequency signal. CW,recv ; (5) Based on the received single-frequency signal frequency f CW,recv Calculate the time-frequency distribution TF of the received FM signal. FM (m FM ,l FM ); In step (2), the energy fluctuation characteristics E of the single-frequency signal are calculated using the following method. CW (m CW Specifically, it includes the following steps: (2-1) Initialize the parameters for calculating the energy fluctuation characteristics of a single-frequency signal, specifically including the initialization of the following parameters: ① Transmit single-frequency signal frequency f CW,trans Initialized to: 0 <f CW,trans <f s / 2 positive real numbers, where f s It is the sampling frequency; ② Transmit single-frequency signal pulse width N CW,trans Initialized to: 0 <N CW,trans Positive integers ≤ N / 2; ③ The single-frequency signal detection window length scaling parameter α is initialized to a positive real number where 0 < α ≤ 1; ④ Single-frequency signal detection window length N CW,window Initialized to: N CW,window =max{4, round{αN CW,trans A positive integer, where max{·} represents the maximum value function and round{·} represents the rounding function; ⑤ Single-frequency signal detection step N CW,step Initialize as: 1≤N CW,step ≤floor{N CW,window A positive integer of size 4, where floor{·} represents the floor function; ⑥ Total number of frames M during single-frequency signal detection CW Initialized to: M CW =floor{(NN CW,window ) / N CW,step Positive integers of size -1; ⑦ Target maximum speed v max Initialize to: 10≤v max Positive real numbers ≤ 20; ⑧ The average underwater sound speed c0 is initialized as a positive real number where 1400 ≤ c0 ≤ 1600; (2-2) Calculate the detection frequency range f of a single-frequency signal CW,det : ； Where, η min and η max The target maximum speed v are respectively max The corresponding minimum and maximum Doppler factors, i.e., η min =(c0-v max ) / (c0+v max ), η max =(c0+v max ) / (c0-v max ); (2-3) Calculate the starting index L of the detection frequency point of the single-frequency signal CW,start The total number of single-frequency signal detection frequency points L CW,total : , ; where f CW,det (1) and f CW,det (2) f represents the detection frequency range of a single-frequency signal. CW,det The first and second elements; (2-4) Extracting the segmented data sequence y of the single-frequency signal detection CW,0 (r CW ),y CW,1 (r CW ),…, ,…, , where the mth CW Segmented data sequence for single-frequency signal detection for: ； in, For single-frequency signal detection segmented time index, For single-frequency signal detection, the frame index is... =0,1,…,M CW -1; (2-5) Calculate the time-frequency distribution TF of the single-frequency signal detection CW (m CW ,l CW ): Among them, l CW This is the frequency index for single-frequency signal detection, where j is the imaginary unit, i.e. , |·| is the modulo function; (2-6) Calculate the energy fluctuation characteristics E of a single-frequency signal CW (m CW ): 。 2. The method for time-frequency analysis of ship active signals based on fine estimation of time-frequency parameters according to claim 1, characterized in that, In step (1), the active signal x(n) of the ship to be processed is obtained using the following method, which specifically includes the following steps: The ship's active signal x(n) to be processed is obtained from real-time data collected from N sampling points by the ship's underwater acoustic sensor, where n = 0, 1, …, N-1; or data from N sampling points is extracted from memory as the ship's active signal x(n), where n = 0, 1, …, N-1, and n is the time index of the ship's active signal x(n) to be processed, and N is the number of signal sampling points, which is N=2. κ κ is a positive integer greater than or equal to 10.

3. The method for time-frequency analysis of ship active signals based on fine estimation of time-frequency parameters according to claim 1, characterized in that, In step (3), the time range n of the received single-frequency signal is calculated using the following method. CW,recv Specifically, it includes the following steps: (3-1) The relative amplitude parameter β for single-frequency signal bandwidth search is initialized to a positive real number where 0 < β < 1; (3-2) Calculate the energy fluctuation characteristics E of a single-frequency signal CW (m CW The time frame index m0 corresponding to the maximum value: ； in, E represents the energy fluctuation characteristics of a single-frequency signal. CW (m CW The time frame index corresponding to the maximum value; (3-3) Calculate the bandwidth amplitude E of a single-frequency signal B : ； (3-4) Search for the energy fluctuation characteristics of a single-frequency signal E CW (m CW The maximum value to the left is less than the bandwidth amplitude E of a single-frequency signal. B Time frame index set m left : ； in, It represents all that satisfy E CW (m CW )≤E B Single-frequency signal detection time frame index m CW The set consisting of max{·} represents the function that takes the maximum value; (3-5) Calculate the frame start index m of a single-frequency signal CW,left : ； in, Denotes the empty set, m left (end) represents the time frame index set m left The last element; (3-6) Search for the energy fluctuation characteristics of a single-frequency signal E CW (m CW The maximum value to the right is less than the bandwidth amplitude E of a single-frequency signal. B Time frame index set m right : ； Where min{·} represents the minimum value function; (3-7) Calculate the frame cutoff index m of a single-frequency signal CW,right : ； Where, m right (1) Represents the time frame index set m right The first element; (3-8) Calculate the time range n of the received single-frequency signal CW,recv : ； in, Indicates search The time index n corresponding to the minimum value, Indicates search The time index n corresponding to the minimum value.

4. The method for time-frequency analysis of ship active signals based on fine estimation of time-frequency parameters according to claim 3, characterized in that, In step (4), the frequency f of the received single-frequency signal is calculated using the following method. CW,recv Specifically, it includes the following steps: (4-1) Initialize the parameters for calculating the frequency of the received single-frequency signal, specifically including the initialization of the following parameters: ① Transmit single-frequency signal period N CW,period Initialized to: N CW,period =round{f s / f CW,trans A positive integer f, where f s This is the sampling frequency, and round{·} represents the rounding function. ② Traverse the set of frequency detectors for pulse width single-frequency signals f CW,trav Initialized as: Its length is N CW,period A vector of all zeros; ③ Traverse the set of peak values ​​of the amplitude spectrum of the pulse width single-frequency signal A CW,trav Initialized as: Its length is N CW,period A vector of all zeros; ④ Traverse the pulse width index d trav Initialized to: d trav =0; (4-2) Calculate the d-th ... trav A single-frequency traversal pulse width signal : ； in, It is the dth trav A time index for traversing a single-frequency pulse width signal. It is the dth trav The number of sampling points for a single-frequency traversal pulse width signal, n CW,recv (1) and n CW,recv (2) Represent the time range n of receiving a single-frequency signal. CW,recv The first and second elements; (4-3) Calculate the d-th ... trav A single-frequency traversal pulse width signal amplitude spectrum : in, It is the dth trav A frequency index for a single-frequency pulse width signal is traversed, where j is the imaginary unit, i.e. , |·| is the modulo function; (4-4) Update the set of frequency detectors f for traversing pulse width single-frequency signals CW,trav The dth trav element f CW,trav (d trav ): ； in, Indicates the search amplitude spectrum Frequency index corresponding to the maximum value ; (4-5) Update the set of peak values ​​of the amplitude spectrum of the pulse width single-frequency signal A. CW,trav The dth trav Element A CW,trav (d trav ): ； Where max{·} represents the function that takes the maximum value; (4-6) Update the traversal pulse width index d trav : ； (4-7) Determine the pulse width index d during traversal trav Does the traversal deadline condition apply? ； If the condition is met, proceed to step (4-8); otherwise, return to step (4-2). (4-8) Calculate the frequency f of the received single-frequency signal. CW,recv : ； Where d0 is the frequency f of the received single-frequency signal. CW,recv The corresponding traversal pulse width peak index, A CW,trav (d trav ) represents the set of peak values ​​of the amplitude spectrum of a pulse-width single-frequency signal. CW,trav The dth trav elements, f CW,trav (d0) represents the set of frequency detectors for traversing pulse width single-frequency signals f. CW,trav The d0th element.

5. The method for time-frequency analysis of ship active signals based on fine estimation of time-frequency parameters according to claim 4, characterized in that, In step (5), the time-frequency distribution TF of the received FM signal is calculated using the following method. FM (m FM ,l FM Specifically, it includes the following steps: (5-1) Initialize the parameters for calculating the time-frequency distribution of the received FM signal, specifically including the initialization of the following parameters: ① Transmit the starting frequency f of the FM signal FM,1,trans Initialized to: 0 <f FM,1,trans <f s / 2 positive real numbers, where f s It is the sampling frequency; ② The cutoff frequency f of the transmitted FM signal FM,2,trans Initialized to: 0 <f FM,2,trans <f s Positive real numbers of 2 / 2; ③ Transmit FM signal pulse width N FM,trans Initialized to: 0 <N FM,trans ≤NN CW,trans Positive real numbers; ④ Time-frequency distribution frequency tolerance parameter λ f Initialize to: 0 < λ f Positive real numbers ≤ 1; ⑤ Time-frequency distribution time tolerance parameter λ t Initialize to: 0 < λ t Positive real numbers ≤ 1; (5-2) Calculate the frequency range f of the received FM signal. FM,recv : ； Where η0=f CW,recv / f CW,trans The target detection Doppler factor, min{·} and max{·} represent the minimum and maximum value functions, respectively; (5-3) Calculate the time range n of the received FM signal FM,recv : ； Where, n CW,recv (1) and n CW,recv (2) Represent the time range n of receiving a single-frequency signal. CW,recv The first and second elements; (5-4) Calculate the time-frequency distribution window length N of the received FM signal. FM,window and the time-frequency distribution step N of the received FM signal FM,step : ； Among them, f FM,recv (1) and f FM,recv (2) represent the frequency range f of the received FM signal respectively. FM,recv The first and second elements, round{·} represents the rounding function, and floor{·} represents the floor function; (5-5) Calculate the starting index L of the received FM signal frequency point FM,start The total number of frequency points L for receiving FM signals FM,total : ； Among them, f FM,range It is the frequency tolerance range for receiving FM signals, f FM,range The first element f FM,range (1)=min{f s / 2, max{0, (1+λ f f FM,recv (1)-λ f f FM,recv (2)}},f FM,range The second element f FM,range (2)=min{f s / 2, (1+λ f f FM,recv (2)-λ f f FM,recv (1)},f FM,recv (1) and f FM,recv (2) represent the frequency range f of the received FM signal respectively. FM,recv The first and second elements; (5-6) Calculate the total number of frames M when the received FM signal time-frequency distribution is calculated. FM : ； (5-7) Extract the segmented data sequence y of the received FM signal FM,0 (r FM ), y FM,1 (r FM ), …, ,…, , where the mth FM Segmented data sequence of received FM signal for: Where, r FM For the time-frequency distribution segmented time index of the received FM signal, m FM For the time-frequency distribution time frame index of the received FM signal, m FM =0,1,…,M FM -1, n FM,recv (1) is the range n of the received FM signal time calculated in step (5-3). FM,recv The first element; (5-8) Calculate and output the time-frequency distribution TF of the received FM signal. FM (m FM , l FM ): Among them, l FM The frequency point index for the time-frequency distribution of the received FM signal, where j is the imaginary unit, i.e. , where |·| is the modulo function.

Images