Tracking method before multi-frame coherent detection of weak target line spectrum

Abstract

A multi-frame coherent detection and tracking method for weak target line spectra is disclosed, based on multi-frame coherent integration. Step 1: Perform multi-frame coherent integration on the passive sonar received data to obtain measured values; Step 2: Establish a state equation including a target presence indicator variable; Step 3: Establish a multi-frame coherent integration measurement equation based on the coherence of the signal phase in the multi-frame data; Step 4: Derive the likelihood ratio function based on multi-frame coherent integration that matches the measurement equation; Step 5: Combine the measured values ​​from Step 1 and the multi-frame coherent detection and tracking model established in Steps 2 to 4, and use a particle filter iterative algorithm to obtain a more robust target line spectrum detection and tracking result. This invention addresses the problem of numerous missed detections in single-frame detection and tracking methods caused by amplitude fluctuations in the received signal in dynamically time-varying marine acoustic channels. It can be applied to most sonar equipment.

Description

Technical Field

[0001] It relates to the field of passive sonar signal processing for target detection and tracking, specifically weak target line spectrum detection. Background Technology

[0002] In underwater acoustic channels, when the line spectrum is continuously observed for a sufficiently long time, the line spectrum amplitude typically exhibits large periodic fluctuations. There are several reasons for this phenomenon. First, the line spectrum components of target radiated noise are mainly generated by mechanical vibrations and the periodic rotation of the propeller. Small changes in these mechanical behaviors lead to instability in the source radiated energy. Second, in the ocean, especially in shallow water environments, the relative motion between the source and receiver causes changes in the multipath interference pattern, resulting in slow, large-period fluctuations in the line spectrum amplitude. Third, random disturbances in the ocean channel, such as surface waves, the movement of inhomogeneous objects in the seawater, and internal waves, cause rapid, small-period fluctuations in the received signal amplitude. These fluctuations in the acoustic signal lead to instability in passive sonar target detection. When the received signal is strong, the target can be easily detected. However, when the signal is weak, the detector may miss the true target or even detect a false target.

[0003] Pre-detection tracking methods can improve the detection performance of passive sonar for weak targets. For example, pre-detection tracking methods based on Hidden Markov Models (HMMs) and Particle Filters (PFFs) are widely used. Compared to HMMs, PFFs are not limited by grids, allowing for more accurate estimation of target frequencies. More importantly, they can simultaneously detect and track multiple line spectra without sequentially removing detected line spectra from the spectrum. However, existing PFFs-based pre-detection tracking methods for line spectra only utilize the amplitude spectrum information of a single or multiple frames after the Discrete Fourier Transform (DFT) of the target line spectrum, ignoring phase information. This leads to discontinuous and unstable tracking results when the target amplitude fluctuates significantly, resulting in numerous missed detections. Summary of the Invention

[0004] To address the problem that existing particle filter-based pre-detection tracking methods for line spectra only utilize the amplitude spectrum information of a single or multiple frames after the Discrete Fourier Transform (DFT) of the target line spectrum, ignoring phase information, which leads to discontinuous and unstable tracking results when the target amplitude fluctuates significantly, resulting in a large number of missed detections, the technical solution provided by this invention is as follows:

[0005] A method for constructing a judgment model before multi-frame coherent detection of weak target line spectra, the method comprising:

[0006] The steps to collect data received by passive sonar and obtain coherence measurement calculation values;

[0007] The steps for acquiring the existence status of the target line spectrum;

[0008] The steps for establishing multi-frame coherent integral measurement equations based on the calculated values ​​and states;

[0009] The step of obtaining the multi-frame coherent likelihood ratio function that satisfies the preset conditions based on the measurement equation is to use it as the model function.

[0010] Furthermore, a preferred embodiment is provided in which the coherence measurement calculation value is specifically obtained based on the data after segmentation and Fourier transform.

[0011] Furthermore, a preferred embodiment is provided in which the existence state is specifically obtained based on the state equation of the target line spectrum as it changes over time.

[0012] Furthermore, a preferred embodiment is provided in which the multi-frame coherent integral measurement equation is specifically established based on a batch processing method.

[0013] Furthermore, a preferred embodiment is provided, wherein the preset condition is specifically: a likelihood ratio function that can effectively determine the existence state of the target under low signal-to-noise ratio.

[0014] Based on the same inventive concept, this invention also provides a device for constructing a judgment model before multi-frame coherent detection of weak target line spectra, the device comprising:

[0015] A module that collects data received by passive sonar and obtains coherence measurement calculation values;

[0016] A module for acquiring the existence status of target line spectra;

[0017] Based on the calculated values ​​and states, a module is established to construct multi-frame coherent integral measurement equations;

[0018] Based on the measurement equation, a multi-frame coherent likelihood ratio function that satisfies preset conditions is obtained, which serves as a module of the model function.

[0019] Based on the same inventive concept, this invention also provides a method for tracking weak target line spectra before multi-frame coherent detection, the method comprising:

[0020] Based on the model established by the method described above, the steps for determining whether the target line spectrum exists are based on the target state.

[0021] Based on the same inventive concept, the present invention also provides a tracking device for multi-frame coherent detection of weak target line spectra, the device comprising:

[0022] Based on the model established by the aforementioned device, a module is used to determine whether the target line spectrum exists according to the target state.

[0023] Based on the same inventive concept, the present invention also provides a computer storage medium for storing a computer program, characterized in that when the computer program is read by a computer, the computer executes the method described thereon.

[0024] Based on the same inventive concept, the present invention also provides a computer, including a processor and a storage medium, characterized in that when the processor reads a computer program stored in the storage medium, the computer executes the method described thereon.

[0025] Compared with the prior art, the advantages of the technical solution provided by the present invention are as follows:

[0026] The weak target line spectrum multi-frame coherent detection pre-tracking method provided by the present invention utilizes the coherent fusion of multi-frame information, which can improve the target signal-to-noise ratio and reduce the influence of target amplitude fluctuations compared with the single-frame detection pre-tracking method, and can effectively improve the line spectrum detection capability.

[0027] The weak target line spectrum multi-frame coherent detection and tracking method provided by the present invention can simultaneously detect and track multiple line spectra, avoiding the combinatorial explosion problem of traditional methods.

[0028] The weak target line spectrum multi-frame coherent detection pre-tracking method provided by this invention, when performing weak target detection with amplitude fluctuations under time-varying ocean channels, utilizes the amplitude and phase information of multi-frame data to establish a dynamic model and measurement model of the multi-frame coherent detection pre-tracking method, and improves the target signal-to-noise ratio through coherent fusion of multi-frame information, thereby improving the detection and tracking performance of targets with amplitude fluctuations.

[0029] The weak target line spectrum multi-frame coherent detection tracking method provided by this invention does not require prior target information, is simple and easy to implement, and can be applied to most sonar equipment. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the batch processing method mentioned in Implementation Method Eleven;

[0031] Figure 2 This is a flowchart illustrating the weak target line spectrum multi-frame coherent detection pre-tracking method mentioned in Implementation Method Eleven;

[0032] Figure 3 This is a schematic diagram of the LOFAR of the received signals obtained by the two methods mentioned in Implementation Method Eleven;

[0033] Figure 4 for Figure 3 A comparative diagram of the line spectrum detection and tracking results. Detailed Implementation

[0034] To make the advantages and benefits of the technical solution provided by the present invention clearer, the technical solution provided by the present invention will now be described in further detail with reference to the accompanying drawings, specifically:

[0035] Implementation Method 1: This implementation method provides a method for constructing a judgment model before multi-frame coherent detection of weak target line spectra. The method includes:

[0036] The steps to collect data received by passive sonar and obtain coherence measurement calculation values;

[0037] The steps for acquiring the existence status of the target line spectrum;

[0038] The steps for establishing multi-frame coherent integral measurement equations based on the calculated values ​​and states;

[0039] The step of obtaining the multi-frame coherent likelihood ratio function that satisfies the preset conditions based on the measurement equation is to use it as the model function.

[0040] Implementation Method 2: This implementation method further defines the method for constructing a judgment model before multi-frame coherent detection of weak target line spectra provided in Implementation Method 1. The coherence measurement calculation value is specifically obtained based on the data after segmentation and Fourier transform.

[0041] Specifically:

[0042] First, the passive sonar observation data is segmented, then the k-th segment of data x k (n) can be represented as:

[0043] x k (n)=s k (n)+g k (n)

[0044] =A k cos[2πf0(n+kM)]+g k (n)

[0045] n=0,1,...,M-1; k=0,1,...,K-1;

[0046] Where s k (n) represents the signal portion of interest in the k-th data segment, i.e., the portion with amplitude A. k A sinusoidal signal with frequency f0; g k (n) represents the noise component in the k-th data segment; n represents the discrete sampling point index, M represents the number of sampling points in each data segment; k represents the segment index, and K represents the total number of data segments (frames). If only positive frequency values ​​are considered, x k The M-point DFT result Y of (n) k (l) can be written as:

[0047]

[0048] Where l represents the discrete frequency unit number, G k (l) represents the noise component g in the k-th data segment. k The M-point DFT result of (n), where j represents the imaginary unit.

[0049] The DFT results of multi-frame data are subjected to DFT transformation at each time step within each discrete frequency cell, the modulus is taken and the result is squared. The maximum value of the result within each discrete frequency cell is taken as the measurement calculation value Z. κ (l) means that coherent integration of multi-frame data is achieved.

[0050] Implementation Method 3: This implementation method is a further limitation on the method for constructing a judgment model before multi-frame coherent detection of weak target line spectra provided in Implementation Method 1. The existence state is specifically obtained based on the state equation of the target line spectrum as it changes over time.

[0051] Specifically,

[0052] Define the state vector at time k as X k =[f k A k ] T f k and A k Let the magnitude and frequency of the target be respectively. Then, the dynamic model of the MFC-TBD method can be established as follows:

[0053] X k+1 =X k +V k ;

[0054] Where V k =[q f ,q A ] T k = 1, 2, ..., K, q f and q A These are random process noises in terms of frequency and amplitude, respectively.

[0055] The target has an indicator variable e k A Markov process that follows a two-state condition, i.e., e k =1 indicates that the target exists, e k =0 represents that the target does not exist, then the probability transition matrix can be defined as:

[0056]

[0057] Where P b P represents the probability of the target birth. d This represents the probability of the target dying.

[0058] Implementation Method 4: This implementation method further defines the method for constructing a judgment model before multi-frame coherent detection of weak target line spectra provided in Implementation Method 1. The multi-frame coherent integral measurement equation is specifically established based on a batch processing method.

[0059] Specifically,

[0060] Let h(·) represent the nonlinear function between the measurements obtained by the passive sonar and the target state, then the measurement model of the MFC-TBD method can be established:

[0061]

[0062] Among them, g k This is random measurement noise. If a target exists at time k, the measurement is entirely determined by the target and random noise. Conversely, if there is no target at time k, the measurement only reflects the characteristics of the noise. By utilizing the difference in the statistical characteristics of the measurement in these two cases, we can infer the state of the target (including its existence) from the measurement.

[0063] In the MFC-TBD method, h(·) is determined by the multi-frame coherent integration process, which is explained in detail below:

[0064] The DFT results from W time points (frames) are merged into a batch, with each batch shifted forward by one frame. Assuming that the line spectrum frequencies and amplitudes within each batch remain unchanged, the κ-th batch can be defined as follows:

[0065]

[0066] Taking advantage of the fact that the phase of the signal within a batch is coherent while the phase of noise is random, the multi-frame coherent integral result can be obtained by performing DFT on the DFT spectral values ​​of multiple data frames within the batch along time intervals and taking the maximum value of the squared magnitude. Therefore, the measurement prediction value of the MFC-TBD method can be defined as:

[0067]

[0068] Where q represents the discrete frequency cell number after performing a DFT along the time interval. The expression is as follows:

[0069]

[0070] Implementation Method 5: This implementation method further defines the method for constructing a judgment model before multi-frame coherent detection of weak target line spectra provided in Implementation Method 1. The preset condition is specifically: a likelihood ratio function that can effectively determine the existence state of the target under low signal-to-noise ratio.

[0071] Based on the multi-frame coherent integration process, the following derivations are made when e k =0 and ek When = 1, the measured value Z κ The probability distribution of (l):

[0072]

[0073]

[0074]

[0075]

[0076] Where I0(·) represents the modified Bessel function of the first kind. p(x) and The expression is as follows:

[0077]

[0078]

[0079] Therefore, the likelihood ratio function can be written as:

[0080]

[0081] The particle filter method provides an effective recursive TBD filtering solution through Monte Carlo simulation. By combining the measured values ​​of the multi-frame coherent integral obtained in step one with the multi-frame coherent detection pre-tracking model established in steps two to four, M particle filters can work in parallel in M ​​discrete frequency units, thereby obtaining the existence probability estimation results of multiple target line spectra.

[0082] Implementation Method Six: This implementation method provides a device for constructing a judgment model before multi-frame coherent detection of weak target line spectra. The device includes:

[0083] A module that collects data received by passive sonar and obtains coherence measurement calculation values;

[0084] A module for acquiring the existence status of target line spectra;

[0085] Based on the calculated values ​​and states, a module is established to construct multi-frame coherent integral measurement equations;

[0086] Based on the measurement equation, a multi-frame coherent likelihood ratio function that satisfies preset conditions is obtained, which serves as a module of the model function.

[0087] Implementation Method Seven: This implementation method provides a tracking method before multi-frame coherent detection of weak target line spectra, the method comprising:

[0088] The model is established based on the method provided in Implementation Method 1, and the step is to determine whether the target line spectrum exists based on the target state.

[0089] Specifically,

[0090] By segmenting the received hydrophone time-domain signal into frames and performing coherent integration on multiple frames, the difference in measurement statistical characteristics is utilized to achieve weak target detection tracking with amplitude fluctuations. This improves the target signal-to-noise ratio, reduces false negatives, and weakens the impact of target amplitude fluctuations, effectively enhancing line spectrum detection capabilities. Specifically, a state model for the Multi-frame Coherent Track-Before-Detect (MFC-TBD) method is first established, ensuring the target state vector includes a target presence indicator variable. Then, based on batch processing techniques, a multi-frame coherent integration measurement equation is established according to the state model and the coherent integration process of multiple frames. A likelihood ratio function based on multi-frame coherent integration, which determines the target's presence state, is derived from the established measurement equation. Finally, the state model, measurement equation, and likelihood ratio function are substituted into a particle filter iterative algorithm to obtain the target's state and existence probability estimation results. The presence of the target line spectrum is then determined based on these results.

[0091] The specific steps include:

[0092] Step 1: After segmenting the passive sonar received data, perform Discrete Fourier Transform (DFT), and then perform coherent integration on the DFT results of multiple segments (or frames) of data to obtain the measurement calculation values.

[0093] Step 2: Establish a state equation for the change of the target line spectrum state (including frequency and amplitude) over time, and describe the target existence state through a target existence indicator variable;

[0094] Step 3: Based on batch processing technology, establish multi-frame coherent integral measurement equations according to the definition of the target state vector in Step 2 and the coherence of the signal phase in the multi-frame data obtained in Step 1.

[0095] Step 4: Based on the measurement equation established in Step 3, derive the likelihood ratio function based on multi-frame coherent integration, which can more effectively determine the existence state of the target under low signal-to-noise ratio.

[0096] Step 5: Combining the measured values ​​of the multi-frame coherent integral obtained in Step 1 and the multi-frame coherent pre-detection tracking model established in Steps 2 to 4, the particle filter iterative algorithm is used to obtain the target's state and existence probability estimation results. Based on the target's state and existence probability estimation results, it is determined whether the target line spectrum exists.

[0097] in,

[0098] After segmenting the passive sonar received data, a DFT transform is performed. The DFT results of multiple frames of data are then transformed at each discrete frequency cell, the modulus is taken, and the square is calculated. The maximum value of the result within each discrete frequency cell is taken as the measurement calculation value Z. κ (l) means that coherent integration of multi-frame data is achieved.

[0099] The state vector of the target line spectrum is defined as X k =[f k A k ], f k and A k These are the target's frequency and amplitude, respectively. The target is defined with an indicator variable e. k A Markov process that follows a two-state condition, i.e., e k =1 indicates that the target exists, e k =0 indicates that the target does not exist. Therefore, the state equation for the target's spectral changes can be established as:

[0100] X k+1 =X k +V k ;

[0101] Where V k =[q f ,q A ] T k = 1, 2, ..., K, q f and q A These are random process noises in terms of frequency and amplitude, respectively.

[0102] The target has an indicator variable e k A Markov process that follows a two-state condition, i.e., e k =1 indicates that the target exists, e k =0 represents that the target does not exist, then the probability transition matrix can be defined as:

[0103]

[0104] Where P b P represents the probability of the target birth. d This represents the probability of the target dying.

[0105] Establish the multi-frame coherent integral measurement equation:

[0106]

[0107] Where h(·) represents the nonlinear function between the measurements obtained by the passive sonar and the target state, g k It is random measurement noise with a variance of σ. 2 x k This represents the received signal at time k.

[0108] If a target exists at time k, the measurement is entirely determined by the target and random noise. Conversely, if there is no target at time k, the measurement only reflects the characteristics of the noise. By utilizing the difference in the statistical characteristics of the measurement in these two cases, we can infer the state of the target (including its existence) from the measurement.

[0109] In the MFC-TBD method, h(·) is determined by the multi-frame coherent integration process, which is explained in detail below:

[0110] First, the observed signal is divided into segments, each containing M sampling points. Then, the k-th segment can be represented as:

[0111] x k (n)=s k (n)+g k (n)

[0112] =A k cos[2πf0(n+kM)]+g k (n)

[0113] n=0,1,...,M-1; k=0,1,...,K-1;

[0114] Where K represents the total number of data frames. When performing Discrete Fourier Transform (DFT) analysis on a signal, leakage effects often occur due to limitations in data analysis length. The DFT discrete unit closest to the true line spectrum frequency f0 is defined as l0. If only positive frequency values ​​are considered, x... k The M-point DFT result of (n) can be written as

[0115]

[0116] h(·) is determined by the multi-frame coherent integration process: the received signal x at time k is... k The M-point DFT result is denoted as Y. k (l), where l represents the discrete frequency unit number. The DFT results of W time points (frames) are merged into a batch, with each batch moving forward one frame. The κ-th batch can then be defined as:

[0117]

[0118] For the DFT results Y of multiple data frames within a batch κ (l),...,Y κ+w-1 (l), by performing DFT on the DFT spectral value corresponding to each l along time and taking the maximum value of its modulus square, the multi-frame coherent integration result can be obtained, that is, the measurement prediction result at time κ:

[0119]

[0120] Where q represents the discrete frequency cell number after performing a DFT along the time interval.

[0121] Based on the established measurement equation, the likelihood functions under the conditions of target presence and non-presence are derived respectively. Dividing the two yields the likelihood ratio function l(Z) based on multi-frame coherent integral. κ (l)|X κ ,e κ ), represented as:

[0122]

[0123] The expression for p(x) is: The expression is I0(·) represents the modified Bessel function of the first kind.

[0124] By combining the measured values ​​of the multi-frame coherent integral obtained in step one with the multi-frame coherent detection pre-tracking model established in steps two to four, the M particle filters are made to work in parallel in the M discrete frequency units, thereby obtaining the existence probability estimation results of multiple target line spectra.

[0125] Implementation Method 8: This implementation method provides a tracking device for multi-frame coherent detection of weak target line spectra before tracking. The device includes:

[0126] A module that determines whether a target line spectrum exists based on the target state, using a model established by the apparatus provided in Implementation Method Six.

[0127] Implementation Method Nine: This implementation method provides a computer storage medium for storing a computer program. When the computer program is read by a computer, the computer executes the method provided in any one of Implementation Methods One to Five.

[0128] Implementation Method 10: This implementation method provides a computer, including a processor and a storage medium, characterized in that when the processor reads a computer program stored in the storage medium, the computer executes the method provided in any one of Implementation Methods 1 to 5.

[0129] Implementation Method Eleven: Combination Figure 1-4 This embodiment describes a preferred implementation of the method provided in Embodiment Seven. Specifically:

[0130] Step 1: First, divide the passive sonar observation data into segments, then the k-th segment of data x k (n) can be represented as:

[0131] x k (n)=s k (n)+g k(n)

[0132] =A k cos[2πf0(n+kM)]+g k (n)

[0133] n=0,1,...,M-1; k=0,1,...,K-1;

[0134] Where s k (n) represents the signal portion of interest in the k-th data segment, i.e., the portion with amplitude A. k A sinusoidal signal with frequency f0; g k (n) represents the additive white noise component in the k-th data segment, which follows a mean of 0 and a variance of σ. 2 The distribution is Gaussian; n represents the discrete sampling point number, M represents the number of sampling points in each data segment; k represents the segment number, and K represents the total number of data segments (frames).

[0135] Perform DFT analysis on the signal, considering only the positive frequency values, x k The M-point DFT result of (n) is denoted as Y. κ (l), where l represents the discrete frequency unit number.

[0136] Step 2: Based on batch processing techniques, perform coherent integration on the DFT results of multiple frames of data. For example... Figure 1 As shown, the DFT results of W time points (frames) are merged into a batch, with each batch shifting forward one frame. Assuming that the line spectrum frequency and amplitude remain unchanged within each batch, then the κth batch... It can be defined as:

[0137]

[0138] Taking advantage of the fact that the phase of the signal within a batch is coherent while the phase of noise is random, the multi-frame coherent integral result can be obtained by performing DFT on the DFT spectral values ​​of multiple data frames within the batch along time intervals and taking the maximum value of the squared magnitude. Therefore, the measurement prediction value of the MFC-TBD method can be defined as:

[0139]

[0140] Where q represents the discrete frequency cell number after performing a DFT along the time interval. The expression is as follows:

[0141]

[0142]

[0143] Step 3: The likelihood ratio function is derived through the multi-frame coherent integration process in Step 2.

[0144] Step 4: Input the measured values ​​of the multi-frame coherent integration obtained above into the particle filter to obtain the estimated value of the target state. Specifically, the state model of the particle filter is the MFC-TBD method's measurement model, along with the likelihood ratio function. M particle filters work collaboratively in M ​​discrete frequency units. By combining the frequency estimate and target existence probability estimate of each discrete frequency unit, we can finally obtain the detection and tracking results for each line spectrum.

[0145] Figure 3 It uses single frames respectively Figure 3 (a) and multiple frames Figure 3 (b) The LOFAR diagram obtained by the coherent integration method. It can be clearly seen from the diagram that the target line spectrum frequency is 100Hz. As the target moves from far to near, the received signal amplitude gradually increases. The target line spectrum is initially submerged by noise, but the signal-to-noise ratio gradually increases over time. It can be clearly seen from the diagram that the multi-frame coherent integration method improves the signal-to-noise ratio of the target line spectrum and can detect the signal earlier.

[0146] Figure 4 Yes Figure 3 The results of the detection and tracking using the LOFAR image show the probability of the target line spectrum presence, with the color bar ranging from 0 to 1. The image shows that the MFC-TBD method in this embodiment can detect the target in approximately 1000 seconds, while the SF-TBD method detects the target later, around 2000 seconds, and the estimated line spectrum presence probability exhibits discontinuity. Therefore, it can be concluded that the MFC-TBD method of this invention improves the target signal-to-noise ratio by coherently integrating multiple frames of data, resulting in superior and more robust target detection and tracking results.

[0147] The above description of several specific embodiments further details the technical solution provided by the present invention in order to highlight the advantages and benefits of the technical solution provided by the present invention. However, the above-described specific embodiments are not intended to limit the present invention. Any reasonable modifications and improvements to the present invention, combinations of embodiments, and equivalent substitutions based on the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0148] In the description of this specification, only preferred embodiments of the present invention are described, and should not be construed as limiting the scope of the invention. Furthermore, the use of terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples" indicates that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or N embodiments or examples. Furthermore, those skilled in the art can combine and integrate the different embodiments or examples described in this specification and the features of different embodiments or examples without contradiction. Additionally, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of the present invention, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified. Any process or method described in the flowcharts or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more N executable instructions for implementing custom logical functions or processes, and the scope of preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order according to the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain. The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, a “computer-readable medium” can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection having one or N wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic device, and portable optical disc read-only memory (CDROM).Furthermore, the computer-readable medium can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory. It should be understood that various parts of the invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0149] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it includes one or a combination of the steps of the method embodiments. Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

Claims

1. A method for constructing a judgment model before multi-frame coherent detection of weak target line spectra, characterized in that, The method includes: The steps to collect data received by passive sonar and obtain coherence measurement calculation values; The steps for acquiring the existence status of the target line spectrum; The steps for establishing multi-frame coherent integral measurement equations based on the calculated values ​​and states; Based on the measurement equation, the multi-frame coherence likelihood ratio function that satisfies the preset conditions is obtained as the model function. The coherence measurement calculation value is specifically obtained based on the data after segmentation and Fourier transform; The multi-frame coherent integral measurement equation is specifically established based on a batch processing method; Specifically, let If we denote the nonlinear function between the measurements obtained by passive sonar and the target state, then the measurement model of the MFC-TBD method can be established: ； in, It is random measurement noise, if If a target is always present, the measurement is entirely determined by the target and random noise; conversely, if... If there is no target at any given time, the measurement only reflects the characteristics of noise. By utilizing the difference in the statistical characteristics of the measurements in the two cases, the target state can be inferred. The part of the signal you are interested in. This indicates that the target exists. This means the target does not exist. yes Time data; In the MFC-TBD method, This is determined by the multi-frame coherent integration process, which is explained in detail below: Will The DFT results at each time step (frame) are merged into a batch, with each batch shifted forward one frame. Assuming that the line spectrum frequencies and amplitudes within each batch remain constant, then the DFT results at each time step (frame) are merged into a batch. Batch definition: ； It is the first Batch DFT results, It is the first Batch DFT results; Taking advantage of the fact that the phase of the signal within a batch is coherent while the phase of noise is random, the multi-frame coherent integral result can be obtained by performing DFT on the DFT spectral values ​​of multiple data frames within the batch along time intervals and taking the maximum value of the squared magnitude. Therefore, the measurement prediction value of the MFC-TBD method can be defined as: ； in This represents the discrete frequency cell number after performing a DFT along the time interval. The expression is as follows: It's the amplitude. It is the discrete frequency unit number obtained after performing an M-point DFT on each frame of signal.

2. The method for constructing a judgment model before multi-frame coherent detection of weak target line spectra according to claim 1, characterized in that, The existence state is specifically obtained based on the state equation of the target line spectrum as it changes over time.

3. The method for constructing a judgment model before multi-frame coherent detection of weak target line spectra according to claim 1, characterized in that, The preset condition is specifically: a likelihood ratio function that can effectively determine the existence state of the target under low signal-to-noise ratio.

4. A device for constructing a judgment model before multi-frame coherent detection of weak target line spectra, characterized in that, The device includes: A module that collects data received by passive sonar and obtains coherence measurement calculation values; A module for acquiring the existence status of target line spectra; Based on the calculated values ​​and states, a module is established to construct multi-frame coherent integral measurement equations; Based on the measurement equation, a multi-frame coherent likelihood ratio function that satisfies the preset conditions is obtained, which serves as a module of the model function. The coherence measurement calculation value is specifically obtained based on the data after segmentation and Fourier transform; The multi-frame coherent integral measurement equation is specifically established based on a batch processing method; Specifically, let If we denote the nonlinear function between the measurements obtained by passive sonar and the target state, then the measurement model of the MFC-TBD method can be established: ； in, It is random measurement noise, if If a target is always present, the measurement is entirely determined by the target and random noise; conversely, if... If there is no target at any given time, the measurement only reflects the characteristics of noise. By utilizing the difference in the statistical characteristics of the measurements in the two cases, the target state can be inferred. The part of the signal you are interested in. This indicates that the target exists. This means the target does not exist. yes Time data; In the MFC-TBD method, This is determined by the multi-frame coherent integration process, which is explained in detail below: Will The DFT results at each time step (frame) are merged into a batch, with each batch shifted forward one frame. Assuming that the line spectrum frequencies and amplitudes within each batch remain constant, then the DFT results at each time step (frame) are merged into a batch. Batch definition: ； It is the first Batch DFT results, It is the first Batch DFT results; Taking advantage of the fact that the phase of the signal within a batch is coherent while the phase of noise is random, the multi-frame coherent integral result can be obtained by performing DFT on the DFT spectral values ​​of multiple data frames within the batch along time intervals and taking the maximum value of the squared magnitude. Therefore, the measurement prediction value of the MFC-TBD method can be defined as: ； in This represents the discrete frequency cell number after performing a DFT along the time interval. The expression is as follows: It's the amplitude. It is the discrete frequency unit number obtained after performing an M-point DFT on each frame of signal.

5. A tracking method before multi-frame coherent detection of weak target line spectra, characterized in that, The method includes: The model established based on the method described in claim 1 includes the step of determining whether the target line spectrum exists based on the target state.

6. A tracking device for multi-frame coherent detection of weak target line spectra before tracking, characterized in that, The device includes: A module for determining whether a target line spectrum exists based on the target state, using a model established by the device described in claim 4.

7. A computer storage medium for storing computer programs, characterized in that, When the computer program is read by the computer, the computer executes the method according to any one of claims 1-3.

8. A computer, comprising a processor and a storage medium, characterized in that, When the processor reads the computer program stored in the storage medium, the computer executes the method according to any one of claims 1-3.

Images