A single vector hydrophone target tracking method and system

Abstract

The invention discloses a single vector hydrophone target tracking method and system, the method comprises the following steps: step 1, a vector hydrophone picks up data, intercepts a low-frequency signal, divides the collected data into M' blocks and samples the data; step 2, performs FFT analysis on each block of collected data, and extracts the instantaneous phase of a multi-frequency unit; step 3, uses an instantaneous phase difference weighting method to obtain a stable estimation value of a target line spectrum and a high-precision estimation value of a target azimuth; step 4, sets a coordinate system, constructs a measurement equation, introduces a linearization operation step through a pseudo-linear method PLE, and converts nonlinearity into a linear approximation; converts the nonlinear problem of the measurement equation into a pseudo-linear problem, and then obtains an analytical solution through a least squares criterion and the azimuth and frequency measurement information obtained in step 3, and obtains an initial state value of the target; step 5, establishes a target state equation, uses a Spherical-Radial criterion, and adopts m volume points with equal weights to realize nonlinear estimation, so as to track underwater targets.

Description

Technical Field

[0001] The invention belongs to the technical field of signal processing, and relates to an underwater target tracking method, and in particular to a high-precision and high-stability single-vector hydrophone target tracking method and system. Background Art

[0002] With the rapid development of modern information science and technology, underwater target tracking theory has attracted the attention of scholars from all over the world, both in the military and civilian fields. Vector hydrophones can simultaneously and co-pointly receive the sound pressure and vibration velocity information of the underwater sound field. The joint processing of sound pressure and vibration velocity can significantly improve the passive sonar direction finding performance. A single vector hydrophone can be used to estimate the direction of the target. In addition, the size and weight of the vector hydrophone are small and easy to deploy. The study of single vector hydrophone passive positioning technology meets the requirements of small hardware scale and can be applied to small working platforms. Therefore, by studying the radiation noise of ship targets received by a single vector hydrophone and then performing signal processing on the received radiation noise, the motion parameters of the target can be solved and underwater target filtering tracking can be performed.

[0003] The document "Analysis of surface target motion based on single vector hydrophone" (Wu Yanqun, Hu Yongming. Acoustic Technology, 2010, 29(04): 361-364) proposes to use a single three-dimensional vector hydrophone to obtain the azimuth and elevation angle of the target, and then use the linear least squares method to linearly fit the target motion trajectory. However, this algorithm is based on batch processing of data, which requires iterative calculation and has a large amount of calculation.

[0004] The document "Research on Target Tracking Method Based on Azimuth-Frequency" (Wang Sicong, Mao Weining. Acoustics and Electronic Engineering, 2004(03):21-24) proposed to obtain the azimuth and frequency information of the target based on the stationary observation station, and use the least squares estimation to obtain an unbiased estimate of the unknown parameters of the target. However, the convergence speed of this method is slow and does not meet the real-time processing requirements of the underwater stationary observation platform.

[0005] The paper "A joint estimation method for target motion parameters by underwater stationary platform" (Liu Shengsong, Chen Shaohua, Chen Chuan. Acta Ordnance, 2016, 37(04): 684-689.) proposed a method to use the estimation result of the least squares estimation algorithm as the iterative initial value of the Kalman filter estimation method, which solved the problem that the iterative initial value has a great influence on the parameter estimation performance. However, the Kalman filter method is simple to calculate and cannot cope with the situation where the target state quantity and the observed quantity have a strong nonlinear relationship. Summary of the invention

[0006] In view of the problems of slow convergence speed, low precision and weak stability of the existing target motion analysis method of single vector hydrophone, the present invention proposes a single vector hydrophone target tracking method and system. The present invention performs phase difference variance weighting on the single-channel line spectrum signal to enhance the line spectrum signal-to-noise ratio, extracts the target line spectrum stable estimation value and the target azimuth high-precision estimation, obtains the target state initial value through pseudo-linear processing, and performs subsequent nonlinear estimation through cubature Kalman filtering. The combined processing method achieves the purpose of highly stable filtering tracking of underwater targets.

[0007] The present invention adopts the following technical scheme:

[0008] A single vector hydrophone target tracking method comprises the following steps:

[0009] Step 1: The vector hydrophone picks up data, intercepts low-frequency signals, divides the collected data into M' blocks and samples and processes them.

[0010] Step 2: Perform FFT analysis on each block of collected data to extract the instantaneous phase of each frequency unit.

[0011] Step 3: Use the instantaneous phase difference weighted method to obtain a stable estimate of the target line spectrum and a high-precision estimate of the target azimuth.

[0012] Step 4: Set the coordinate system, determine the relationship between the target azimuth and its frequency and the target position and speed based on the geometric relationship, construct the measurement equation, introduce the linearization operation steps through the pseudo-linear method (PLE), transform the nonlinear system into an approximation of the linear system, and then use the linear system method for processing and analysis. The nonlinear problem of the measurement equation is transformed into a pseudo-linear problem, and then the analytical solution is obtained through the least squares criterion and the azimuth and frequency measurement information obtained in step 3 to obtain the initial state value of the target.

[0013] Step 5, establish the target state equation, use the azimuth and frequency information obtained in step 3 and the analytical solution obtained in step 4 as the initial value input of the cubature Kalman filter (CKF) algorithm, use the spherical-radial criterion, and use m equally weighted volume points to realize nonlinear estimation, so as to achieve the effect of high stability and high precision filtering tracking of underwater targets.

[0014] Preferably, in step 1, low-frequency information is extracted from the received signal at a sampling rate of f s The received signal is sampled to obtain the discrete signal p(n'T s ),v x (n'T s ),v y (n'Ts ).

[0015] Preferably, in step 2, the obtained discrete signal is subjected to fast Fourier transform analysis according to formula (1), and the discrete signal of the sound pressure channel p is p(n'T s ), extract the analysis result P(kω s )

[0016]

[0017] Among them, ω s =2π / KT s is the frequency sampling interval in FFT analysis;

[0018] Set the vibration velocity channel v x The discrete signal v x (n'T s ) to perform fast Fourier transform analysis and extract the analysis result Vx(kω s )

[0019]

[0020] Set the vibration velocity channel v y The discrete signal v y (n'T s ) to perform fast Fourier transform analysis and extract the analysis result Vy(kω s )

[0021]

[0022] Preferably, in step 3, the instantaneous phase difference of all frequency units of all collected data blocks is calculated according to formula (4), the sound pressure channel is p, and the instantaneous phase difference variance value δ is obtained p (kω s ):

[0023]

[0024] in, is the average value of the instantaneous phase difference of all frequency units.

[0025] Vibration velocity channel v x The instantaneous phase difference variance value δx(kωs) obtained is:

[0026]

[0027] Vibration velocity channel v yThe instantaneous phase difference variance value δy(kωs) obtained is:

[0028]

[0029] The instantaneous phase difference of the channel is After mean filtering, we get The variance weighted statistics of the instantaneous phase difference of the frequency units under the three channels are performed according to formula (7) to obtain the final spectrum and signal line spectrum estimation values:

[0030]

[0031] After the instantaneous phase difference weighting processing Perform conjugate cross-spectroscopy separately:

[0032]

[0033] Calculate the target direction at each frequency, take the average of all frequencies to get the azimuth at the current time t

[0034]

[0035] Preferably, in step 4, based on the geometric position relationship between the vector hydrophone and the target radiation source, it is assumed that the state equation of a single vector hydrophone is: The target's motion state is uniform linear motion, and the state vector is Establish the target measurement equation:

[0036]

[0037] Where c is the speed of sound wave propagation, and the target radiation source frequency f c Added to the target state vector and estimated as the target motion parameter, the initial state vector of the target is:

[0038]

[0039] In order to obtain the initial state solution of the target, the orientation and frequency of the measurement equation are processed pseudo-linearly to obtain a linear equation about the target state parameters, which is sorted and simplified into a matrix form:

[0040] Z g =A g Y gt +PV g (12)

[0041] Through the cost function of the target:

[0042]

[0043] Get the target state Y t The weighted least squares solution of :

[0044]

[0045] Preferably, in step 5, the state equation of the target is established as shown in formula (15):

[0046] X t+1 =ΦX t +Gη t (15)

[0047] Based on the target’s position, frequency information and the target’s initial state estimation value, the spherical-radial criterion of the cubature Kalman filter algorithm is used to select m equally weighted volume points (ξ j ,ω j ) to realize nonlinear estimation, where ξ j is the volume point, ω j The weight corresponding to 1 / m is used to solve the motion parameters of the target and perform underwater target tracking.

[0048] The present invention also discloses a single vector hydrophone target tracking system, which is based on the above method and includes the following modules:

[0049] Data acquisition module: vector hydrophone picks up data, intercepts low-frequency signals, divides the collected data into M blocks and samples and processes them;

[0050] Instantaneous phase extraction module: performs FFT analysis on each block of collected data to extract the instantaneous phase of the multi-frequency unit;

[0051] Estimation module: uses the instantaneous phase difference weighting method to obtain a stable estimate of the target line spectrum and a high-precision estimate of the target azimuth;

[0052] Target initial state value acquisition module: Set the coordinate system, determine the relationship between the target azimuth and frequency and the target position and speed based on the geometric relationship, construct the measurement equation, introduce the linearization operation steps through the pseudo-linear method PLE, and convert the nonlinearity into a linear approximation; convert the nonlinear problem of the measurement equation into a pseudo-linear problem, and then obtain the analytical solution through the least squares criterion and the azimuth and frequency measurement information obtained by the valuation module to obtain the initial state value of the target;

[0053] Tracking module: establish the target state equation, take the azimuth and frequency information obtained by the estimation module and the initial solution obtained by the target initial state value acquisition module as the input of the cubature Kalman filter algorithm CKF, use the Spherical-Radial criterion and m equal-weighted cubature points to realize nonlinear estimation to track underwater targets.

[0054] In summary, the present invention discloses a single vector hydrophone target tracking method and system. The present invention uses a single vector hydrophone to detect underwater sound field information, and through the joint information processing of sound pressure and particle vibration velocity, the instantaneous phase difference variance of each channel frequency unit is weighted, the background noise is suppressed, the line spectrum signal-to-noise ratio is enhanced, and the target line spectrum stable estimation value and the target azimuth high-precision estimation value are extracted, and then the azimuth-frequency multi-dimensional information joint tracking and pseudo-linear batch data processing are performed to obtain the weighted least squares solution of the target data, and the initial value of the target's motion state is obtained. Finally, the obtained initial value result is used as the input of the cubature Kalman filter tracking algorithm for high-precision processing. The present invention obtains the target azimuth-frequency multi-dimensional information through a single vector hydrophone, and organically combines the analytical nature of the pseudo-linear method, that is, the characteristic of obtaining a simple initial solution, with the advantages of high estimation accuracy, not easy to diverge, and small amount of calculation of the cubature Kalman filter, so as to achieve high stability and high-precision tracking of underwater targets. The present invention has the advantages of no initial value, fast convergence speed, high accuracy, and good stability. The simulation results show that the target state predicted by the joint tracking algorithm of the present invention is highly consistent with the target's actual state, which proves the effectiveness and good engineering practicality of the processing method of the present invention.

[0055] The present invention has the following beneficial effects:

[0056] 1. The present invention is based on the cross-spectrum azimuth estimation technology of vector hydrophone, which makes full use of the acoustic vector sensor to pick up more comprehensive sound field information, and uses the sound pressure and particle vibration velocity joint information processing technology with strong line spectrum detection capability to obtain the target's azimuth and frequency information.

[0057] 2. The present invention is applicable to targets that maintain uniform linear motion and have characteristic line spectra within the working frequency band of the hydrophone, and can fit the target track according to its azimuth and frequency information.

[0058] 3. The joint tracking algorithm involved in the present invention does not need to obtain the initial state value of the target, and the algorithm has fast convergence speed, high accuracy and is relatively stable. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 The present invention is a flowchart of a single vector hydrophone target tracking method according to a preferred embodiment of the present invention.

[0060] Figure 2 This is a diagram showing the geometric position relationship between the vector hydrophone and the target in the present invention.

[0061] Figure 3 This is the effect diagram of the PLE algorithm in the present invention.

[0062] Figure 4 This is a diagram of target frequency variation in the simulation of the present invention.

[0063] Figure 5This is a graph of the trajectory results tracked by the patented joint algorithm in the simulation of the present invention.

[0064] Figure 6 This is a local enlarged view of the trajectory tracking in the simulation of the present invention.

[0065] Figure 7 This is a target position error analysis diagram in the simulation of the present invention.

[0066] Figure 8 This is a target speed error analysis diagram in the simulation of the present invention.

[0067] Fig. 9 The figure is a block diagram of a single vector hydrophone target tracking system according to a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0068] The preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0069] The present invention relates to a single vector hydrophone target tracking method and system. The present invention receives the target signal through a single vector hydrophone, utilizes the relatively stable characteristic of the instantaneous phase difference corresponding to the target line spectrum frequency unit, performs instantaneous phase difference variance weighting on each channel frequency unit, extracts the target's azimuth and frequency stable estimation value, and organically combines the analytical nature of the pseudo-linear method with the advantages of high estimation accuracy, low divergence, and small amount of calculation of the cubature Kalman filter, thereby achieving highly stable tracking of underwater targets. The joint tracking algorithm of the present invention does not require the initial state information of the target, and has the advantages of fast convergence speed, high accuracy, and good stability.

[0070] like Figure 1 As shown, a single vector hydrophone target tracking method according to a preferred embodiment of the present invention has the following detailed steps:

[0071] Step 1: The two-dimensional vector hydrophone picks up data, intercepts the low-frequency signal, divides the collected data into M blocks, and samples the received signal with a sampling rate of f s , and obtain the discrete signals p(n'T s ),v x (n'T s ),v y (n'T s ), where T s =1 / f s , n'=1,2,···,N', N' is the data length.

[0072] Step 2: Perform fast Fourier transform analysis on the discrete signal obtained in step 1, and use the discrete signal p(n'T s ) as an example, the analysis result P(kω s ) is the instantaneous phase of each frequency unit k=1,2,···,K,m'=1,2,···,M',where K is the number of frequency units obtained by FFT analysis of each block of collected data, and M' is the number of blocks of collected data:

[0073]

[0074] Among them, ω s =2π / KT s is the frequency sampling interval in FFT analysis.

[0075] Set the vibration velocity channel v x The discrete signal v x (n'T s ) to perform fast Fourier transform analysis and extract the analysis result Vx(kω s )

[0076]

[0077] Set the vibration velocity channel v y The discrete signal v y (n'T s ) to perform fast Fourier transform analysis and extract the analysis result Vy(kω s )

[0078]

[0079] Step 3: Taking the sound pressure channel p as an example, the variance of the instantaneous phase difference of all frequency units of all collected data blocks is calculated to obtain the instantaneous phase difference variance value δ p (kω s ):

[0080]

[0081] in, is the average value of the instantaneous phase difference of all frequency units.

[0082] Then the vibration velocity channel v x The instantaneous phase difference variance value δx(kωs) obtained is:

[0083]

[0084] Then the vibration velocity channel v y The instantaneous phase difference variance value δ is obtained y (kω s ):

[0085]

[0086] Then the corresponding channel instantaneous phase difference is After mean filtering, we get

[0087] The variance-weighted statistics of the instantaneous phase difference of the frequency units under the three channels are performed to obtain the final spectrum and signal line spectrum estimation values:

[0088]

[0089] After the instantaneous phase difference weighting processing Perform conjugate cross-spectrum respectively and take its real part:

[0090]

[0091] Calculate the target direction at each frequency, take the average of all frequencies to get the azimuth at the current time t

[0092]

[0093] Step 4: Establish the target measurement equation. The geometric position relationship between the vector hydrophone and the target radiation source is as follows: Figure 2 As shown, at the sampling time t, the state equation of the single vector hydrophone is The target's motion state is uniform linear motion, and the state vector is where x gt and gt Respectively represent the position coordinates of the target in the X and Y directions, and Represent the speed of the target in the X and Y directions respectively. Then the geometric relationship between the observer and the target can be obtained:

[0094]

[0095]

[0096] When there is relative motion between the target and the vector hydrophone, the vector hydrophone will produce a Doppler frequency shift when receiving the target radiation frequency. t will contain the Doppler frequency f dk Assume that the radiation source frequency f during the observation time c Remain unchanged, that is:

[0097]

[0098] Where c is the speed of sound wave propagation.

[0099] Then the measurement equation of the target is:

[0100]

[0101] The target radiation source frequency f c Added to the target state vector, it can be used as the target motion parameter estimation, then the target initial state vector is:

[0102]

[0103] In order to obtain the initial state solution of the target, the orientation and frequency of the measurement equation are pseudo-linearly processed to obtain a linear equation about the target state parameter. Substituting the state vector into equations (10) and (11), we can obtain:

[0104] x ot cosθ t -y ot sinθ t =[cosβ t -sinθ t (t-1)T cosθ t -(t-1)T sinθ t 0]Y gt (15)

[0105]

[0106] Then the noisy measurement is:

[0107]

[0108]

[0109] Among them, v θt and v ft The normal random noises are expressed as the measurement direction and the measurement frequency respectively, then the measurement equation of the system can be expressed as:

[0110]

[0111] Among them, r t is the distance from the target to the vector hydrophone, since f c and r t It is an unknown quantity, so it can be replaced by the frequency measurement mean and distance estimation during the solution process.

[0112] The above formula is simplified into a matrix form, and when N observations are obtained, all observations are batch processed. The measurement equation at this time is:

[0113] Z g =A g Y gt +PVg (20)

[0114] in,

[0115] Then the cost function of the target is:

[0116]

[0117] Among them, N g is the covariance matrix of azimuth and frequency measurement errors.

[0118] By taking the derivative of the above cost function, we can get the target state Y t The weighted least squares solution of :

[0119]

[0120] Figure 4 This is the simulation result diagram of the pure PLE algorithm tracking the target.

[0121] Step 5: Establish the state equation of the target. When the target and the observation platform are in the same plane, the vector hydrophone is fixed at (x ot ,y ot ), the target is at (x gt ,y gt ) is imposed at a speed For uniform linear motion, the state equation of the target can be expressed as:

[0122] X t+1 =ΦX t +Gη t (twenty two)

[0123] in, is the state vector of the target at time t, is the state transfer matrix, T is the observation period; η t is the covariance matrix Q η Two-dimensional zero-mean Gaussian white noise, is the state noise transfer matrix, so the state noise covariance matrix is ​​Q t =E[(Gη t )(Gη t ) T ]=GQ η G T .

[0124] According to the above state equation, the measurement equation in step 4, and the azimuth and frequency information obtained in step 3 are substituted into the cubature Kalman filter.

[0125] The core idea of ​​the volumetric Kalman filter algorithm is to use the Spherical-Radial criterion and adopt m equally weighted volume points (ξ j ,ω j ) to realize nonlinear estimation, where ξ j is the volume point, ω j is the weight corresponding to 1 / m, and the algorithm flow is as follows:

[0126] 1) Time update

[0127] ① Calculate volume points

[0128]

[0129] Among them, S k-1 =chlo{P k-1}, P k-1 is the covariance matrix, chol{} represents the Cholesky decomposition of the matrix, that is,

[0130] ② Calculate the volume points propagated through the state equation

[0131]

[0132] ③State prediction

[0133]

[0134] ④ Covariance prediction

[0135]

[0136] 2) Measurement update

[0137] ① Calculate volume points

[0138]

[0139] Among them, S k|k-1 =chol{P k|k-1}.

[0140] ② Calculate the volume points propagated through the measurement equation

[0141] Z j,k|k-1 =h(X j,k|k-1 ) (28)

[0142] ③Measurement prediction

[0143]

[0144] ④ Estimation of the new information covariance

[0145]

[0146] ⑤ Estimation of mutual covariance

[0147]

[0148] ⑥Calculate Kalman gain

[0149]

[0150] ⑦Status Update

[0151]

[0152] ⑧Estimated covariance

[0153]

[0154] Performance Analysis:

[0155] In order to verify the stability and accuracy of the pseudo-linear estimation-volumetric Kalman joint tracking algorithm described in the present invention, the simulation analysis results are given below. Assume that the single vector hydrophone is fixed at the coordinates (-500m, 0m), and the target radiation source moves at a uniform linear speed of (15m / s, 15m / s) from (-1000m, 1000m). The observation period is 1s, and the observation time lasts for 1000s. The joint tracking algorithm described in the present invention is used for processing, of which the first 20s are pseudo-linear processing to obtain the weighted least squares solution of the target, and the next 980s are volumetric Kalman filter processing using the result obtained by the pseudo-linear estimation algorithm as the initial value. Assume that the target radiation source maintains f during the observation time. c =1000Hz unchanged, the measurement error of angle and frequency is R k =diag[1 2 ,0.5 2 ].

[0156] The present invention uses the root mean square error (RMSE) as an evaluation index to describe the tracking performance of the algorithm, which is defined as:

[0157]

[0158] Among them, (x i ,y i ) represents the actual position of the target at the i-th moment, Represents the predicted value of the target position at the i-th moment.

[0159] Using 5000 Monte Carlo simulations, Figure 4 is a frequency variation diagram received by a single vector hydrophone. The trajectory prediction results of the joint tracking algorithm and the PLE algorithm of the present invention are as follows: Figure 5 ,6 As shown in Figure 2, the positioning and tracking performance is as follows: Figure 7 , 8 As shown. From the results, it can be seen that the joint tracking algorithm converges in about 120s, the PLE algorithm converges in about 350s, and after convergence, the subsequent position tracking effect of the joint tracking algorithm is relatively stable, while the PLE algorithm fluctuates near the convergence value. Compared with the PLE algorithm, the position estimation accuracy of the joint tracking algorithm of the present invention is improved by 21.29%, and the speed estimation accuracy is improved by 22.83%. The average single-time consumption time of the joint tracking algorithm of the present invention is 0.05s, and the average single-time consumption time of the PLE algorithm is 115.64s.

[0160] like Fig. 9 As shown, the embodiment of the present invention discloses a single vector hydrophone target tracking system, which includes the following modules based on the above method embodiment:

[0161] Data acquisition module: vector hydrophone picks up data, intercepts low-frequency signals, divides the collected data into M blocks and samples and processes them;

[0162] Instantaneous phase extraction module: performs FFT analysis on each block of collected data to extract the instantaneous phase of the multi-frequency unit;

[0163] Estimation module: uses the instantaneous phase difference weighting method to obtain a stable estimate of the target line spectrum and a high-precision estimate of the target azimuth;

[0164] Target initial state value acquisition module: Set the coordinate system, determine the relationship between the target azimuth and frequency and the target position and speed based on the geometric relationship, construct the measurement equation, introduce the linearization operation steps through the pseudo-linear method PLE, and convert the nonlinearity into a linear approximation; convert the nonlinear problem of the measurement equation into a pseudo-linear problem, and then obtain the analytical solution through the least squares criterion and the azimuth and frequency measurement information obtained by the valuation module to obtain the initial state value of the target;

[0165] Tracking module: establish the target state equation, take the azimuth and frequency information obtained by the estimation module and the initial solution obtained by the target initial state value acquisition module as the input of the cubature Kalman filter algorithm CKF, use the Spherical-Radial criterion and m equal-weighted cubature points to realize nonlinear estimation to track underwater targets.

[0166] For other contents of this embodiment, please refer to the above method embodiment.

[0167] In summary, the present invention discloses a single-vector hydrophone target tracking method and system. The present invention utilizes a single-vector hydrophone to detect underwater sound field information, and through the sound pressure and particle vibration velocity joint information processing technology, performs instantaneous phase difference variance weighting on each channel frequency unit, suppresses background noise, enhances the line spectrum signal-to-noise ratio, extracts the target line spectrum stable estimation value and the target azimuth high-precision estimation value, and then organically combines the analytical property of the pseudo-linear method, that is, the characteristic of obtaining a simple initial solution, with the advantages of high estimation accuracy, not easy to diverge, and small amount of calculation of the volumetric Kalman filter, and finally realizes highly stable and high-precision tracking of underwater targets. The present invention has the advantages of no need for initial values, fast convergence speed, high accuracy, and good stability. The simulation results show that the consistency between the target state predicted by the joint tracking algorithm described in the present invention and the actual state of the target is high, which proves the effectiveness and good engineering practicality of the processing method of the present invention.

[0168] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. A person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the scope of the present application should be based on the protection scope of the claims.

Claims

1. A single vector hydrophone target tracking method, characterized in that The following steps are involved: Step 1: The vector hydrophone picks up data, intercepts low-frequency signals, divides the collected data into M' blocks and samples and processes them; Step 2: Perform FFT analysis on each block of collected data to extract the instantaneous phase of the multi-frequency unit; Step 3: Use the instantaneous phase difference weighting method to obtain a stable estimate of the target line spectrum and a high-precision estimate of the target azimuth; Step 4: Set the coordinate system, determine the relationship between the target azimuth and frequency and the target position and speed based on the geometric relationship, construct the measurement equation, introduce the linearization operation steps through the pseudo-linear method PLE, and convert the nonlinearity into a linear approximation; convert the nonlinear problem of the measurement equation into a pseudo-linear problem, and then obtain the analytical solution through the least squares criterion and the azimuth and frequency measurement information obtained in step 3 to obtain the initial state value of the target; Step 5: Establish the target state equation, use the azimuth and frequency information obtained in step 3 and the initial solution obtained in step 4 as the input of the cubature Kalman filter algorithm CKF, use the Spherical-Radial criterion, and use m equally weighted cubature points to realize nonlinear estimation to track underwater targets.

2. The single vector hydrophone target tracking method according to claim 1, characterized in that: In step 1, the low-frequency information of the received signal is extracted at a sampling rate of f s The received signal is sampled to obtain a discrete signal p(n'T s ), vibration velocity channel v x The discrete signal v x (n'T s ) and the vibration velocity channel v y The discrete signal v y (n'T s ), where T s =1 / f s , n'=1,2,···,N', N' is the data length.

3. The single vector hydrophone target tracking method according to claim 2, characterized in that: In step 2, the obtained discrete signal is subjected to fast Fourier transform analysis according to formula (1), and the discrete signal of the sound pressure channel p is p(n'T s ), extract the analysis result P(kω s ) Among them, ω s =2π / KT s is the frequency sampling interval in FFT analysis, k=1,2,···,K, m'=1,2,···,M', K is the number of frequency units obtained by FFT analysis of each block of acquired data, and M' is the number of blocks of acquired data; Set the vibration velocity channel v x The discrete signal v x (n'T s ) to perform fast Fourier transform analysis and extract the analysis result Vx(kω s ) Set the vibration velocity channel v y The discrete signal v y (n'T s ) Perform fast Fourier transform analysis and extract the analysis result V y (kω s ) 4. The single vector hydrophone target tracking method according to claim 3, characterized in that: In step 3, the variance of the instantaneous phase difference of all frequency units of all collected data blocks is calculated according to formula (4), with the sound pressure channel being p, and the instantaneous phase difference variance value δ is obtained p (kω s ): in, is the average value of the instantaneous phase difference of all frequency units; Vibration velocity channel v x The instantaneous phase difference variance value δ is obtained x (kω s ): Vibration velocity channel v y The instantaneous phase difference variance value δ is obtained y (kω s ): The instantaneous phase difference of the channel is After mean filtering, we get The variance weighted statistics of the instantaneous phase difference of the frequency units under the three channels are performed according to formula (7) to obtain the final spectrum and signal line spectrum estimation values: After the instantaneous phase difference weighting processing Perform conjugate cross-spectroscopy separately: Calculate the target direction at each frequency, take the average of all frequencies to get the azimuth at the current time t 5. The single vector hydrophone target tracking method according to claim 4, characterized in that: In step 4, based on the geometric position relationship between the vector hydrophone and the target radiation source, the state equation of the single vector hydrophone is assumed to be: The target's motion state is uniform linear motion, and the state vector is Establish the target measurement equation: Where c is the speed of sound wave propagation, and the target radiation source frequency f c Added to the target state vector and estimated as the target motion parameter, the initial state vector of the target is: The orientation and frequency of the measurement equation are processed pseudo-linearly to obtain a linear equation about the target state parameter, which is written in matrix form as follows: Z g =A g AND gt +PV g (12) in, R(X t )=diag(r t ), r t is the distance from the target to the vector hydrophone; Through the cost function of the target: Among them, N g is the covariance matrix of azimuth and frequency measurement errors; Get the target state Y t The weighted least squares solution of :

6. The single vector hydrophone target tracking method according to claim 5, characterized in that: In step 5, the state equation of the target is established as shown in equation (15): X t+1 =ΦX t +Gη t (15) in, is the state vector of the target at time t, is the state transfer matrix, T is the observation period; η t is the two-dimensional zero-mean Gaussian white noise with a covariance matrix, is the state noise transfer matrix; Based on the target’s orientation, frequency information and the target’s initial state estimation value, the spherical-radial criterion of the cubature Kalman filter algorithm is used to select m equally weighted cubature points (ξ j ,ω j ) to realize nonlinear estimation, where ξ j is the volume point, ω j The weight corresponding to 1 / m is used to solve the target's motion parameters and perform underwater target tracking.

7. A single vector hydrophone target tracking system, based on the method according to any one of claims 1 to 6, characterized in that Includes the following modules: Data acquisition module: vector hydrophone acquires data, intercepts low-frequency signals, divides the acquired data into M' blocks and samples and processes them; Instantaneous phase extraction module: performs FFT analysis on each block of collected data to extract the instantaneous phase of the multi-frequency unit; Estimation module: uses the instantaneous phase difference weighting method to obtain a stable estimate of the target line spectrum and a high-precision estimate of the target azimuth; Target initial state value acquisition module: Set the coordinate system, determine the relationship between the target azimuth and frequency and the target position and speed based on the geometric relationship, construct the measurement equation, introduce the linearization operation steps through the pseudo-linear method PLE, and convert the nonlinearity into a linear approximation; convert the nonlinear problem of the measurement equation into a pseudo-linear problem, and then obtain the analytical solution through the least squares criterion and the azimuth and frequency measurement information obtained by the valuation module to obtain the initial state value of the target; Tracking module: establish the target state equation, take the azimuth and frequency information obtained by the estimation module and the initial solution obtained by the target initial state value acquisition module as the input of the cubature Kalman filter algorithm CKF, use the Spherical-Radial criterion and m equal-weighted cubature points to realize nonlinear estimation to track underwater targets.

Images