Underwater acoustic navigation enhancement method, system, storage medium and device

Abstract

The present invention belongs to the field of underwater acoustic navigation technology and provides an underwater acoustic navigation enhancement method, system, storage medium, and device. The method performs standardized preprocessing on acquired sonar data and seawater parameter data, blocks the preprocessed data, constructs and solves a time observation equation taking into account the propagation time of the sound waves, and jointly estimates the azimuth seabed beacon position and the nadir total delay error parameter. The estimated nadir total delay error parameter and its corresponding time series are used as a training sample set, and a nadir total delay error prediction model is constructed and trained using the least squares support vector machine method. The model uses the sum of the double absolute values ​​of all error variables and weight vectors as the objective function. The trained model is used to correct the time-varying sound velocity error in long-baseline underwater acoustic navigation. Based on the corrected time-varying sound velocity error, an extended Kalman filter is used for navigation solution. The present invention improves the accuracy of underwater acoustic navigation.

Description

Technical Field

[0001] The present invention belongs to the technical field of underwater acoustic navigation, and in particular relates to an underwater acoustic navigation enhancement method, system, storage medium and equipment. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Marine survey research, resource development, and utilization require the support of high-precision, highly reliable marine geodetic benchmarks and marine navigation technologies. Because electromagnetic waves are severely attenuated when propagating underwater, acoustic waves are chosen as the primary information carrier for underwater positioning and navigation. The long base line (LBL) underwater acoustic navigation system, consisting of a surface unit, a Global Navigation Satellite System (GNSS) antenna, and a seafloor reference network, offers a wide range and high navigation accuracy, meeting the navigation accuracy requirements of most civil and military applications. However, due to the complex ocean environment, underwater acoustic navigation positioning is affected by multiple errors, such as sound velocity error and time delay; among these, sound velocity error has a greater impact on navigation results. Therefore, suppressing the influence of sound velocity error is key to further improving underwater navigation positioning accuracy.

[0004] To mitigate the impact of sound velocity errors, scholars in related fields have conducted extensive research. While existing methods can somewhat mitigate the impact of sound velocity errors on LBL underwater acoustic navigation, various challenges remain. Because sound velocity is influenced by multiple environmental factors, such as depth, temperature, and salinity, and exhibits temporal and spatial variations, an accurate observation model combined with an effective parameter estimation algorithm can effectively mitigate the impact of systematic errors in acoustic positioning. Current underwater acoustic navigation methods, such as ray tracing and inversion methods, rely on precise modeling of the ocean environment or complex physical assumptions, making them difficult to adapt to the spatiotemporal uncertainties and nonlinear variations of the sound velocity field. Furthermore, traditional filtering algorithms are computationally complex and lack real-time performance, and most error compensation models rely on redundant observation data, limiting their applicability when underwater observation information is limited. In particular, the Nadir Total Delay (NTD) error, which reflects the time-varying nature of sound velocity and is primarily used in ocean benchmark calibration, is a particularly important factor. However, existing methods often use linear or deterministic models to accurately capture its complex dynamic characteristics. Traditional underwater acoustic navigation models often ignore the impact of systematic errors or simply treat them as white noise, resulting in inaccurate results. Summary of the Invention

[0005] In order to solve the above problems, the present invention proposes an underwater acoustic navigation enhancement method, system, storage medium and device. The present invention can significantly improve the accuracy of LBL underwater acoustic navigation and provide a new idea for high-precision underwater positioning in complex marine environments.

[0006] According to some embodiments, the present invention adopts the following technical solutions:

[0007] A method for enhancing underwater acoustic navigation comprises the following steps:

[0008] The acquired sonar data and seawater parameter data are standardized and preprocessed. The preprocessed data are then divided into blocks, each corresponding to a nadir total delay parameter. Taking into account the propagation time of the sound wave, a time observation equation is constructed and solved to jointly estimate the azimuth seabed beacon position and the nadir total delay error parameter.

[0009] The estimated nadir total delay error parameters and their corresponding time series are used as training sample sets, and a least squares support vector machine method is used to construct and train a nadir total delay error prediction model. The nadir total delay error prediction model takes minimizing the sum of the double absolute values ​​of all error variables and weight vectors as the objective function;

[0010] The trained nadir total delay error prediction model is used to correct the time-varying error of sound speed in long baseline underwater acoustic navigation, and an extended Kalman filter is used to perform navigation solution based on the corrected time-varying error of sound speed.

[0011] As an optional implementation, the process of segmenting the preprocessed data includes: dividing each set number of data points into an independent data block, each data block corresponds to a nadir total delay parameter, and assuming that the nadir total delay parameter remains constant within a set time period, that is, within this time period, the time-varying characteristics of the sound speed are considered to be uniformly distributed.

[0012] As an optional embodiment, the process of constructing the time observation equation includes: taking into account the first i The propagation time of the kth transmission of a transponder is expressed as the time observation equation using the nadir total delay NTD parameter:

[0013] ;

[0014] ;

[0015] in, and are the observed and calculated round trip times, respectively, represents the position of the i-th seabed transponder, represents the position of the shipborne transducer, represents the lever arm offset, and Respectively represent the observation time of the shipborne transducer when transmitting and receiving, and the reference sound speed profile Obtained by sound velocity measuring instrument, is the travel time delay integrated along the lowest point of the platform, corresponding to the defined nadir delay parameter, Corresponding to ,coefficient is a scaling factor determined by the angle of incidence of the acoustic ray path Sure.

[0016] As an optional implementation, the process of solving the time observation equation includes: for all observation data in a data block, constructing a simplified residual equation:

[0017] ;

[0018] Arrange the equations for all observation data into a linear system:

[0019] ;

[0020] In the above formula, X is the unknown parameter vector, which includes the seabed reference point coordinate correction and the nadir total delay NTD parameter of each data block; A is the design matrix; LS is the observation vector;

[0021] Solve using weighted least squares method:

[0022] ;

[0023] In the above formula, P Is the weight matrix, which can be the unit matrix initially;

[0024] Update through iteration X , until X If the maximum absolute value of is less than a predetermined threshold, the NTD error estimation value can be obtained.

[0025] As an optional implementation, the process of constructing a nadir total delay error prediction model using a least squares support vector machine method includes: constructing an objective function based on a least squares support vector machine model represented in a feature space:

[0026] ;

[0027] The above objective function is subject to the equality constraints:

[0028] ;

[0029] in, Represents the model output data, is the weight vector, Is a mapping function, which represents the mapping function from input space to feature space; b represents the model bias parameter, represents the error variable; is the regularization parameter;

[0030] Then introduce the Lagrangian factor to the constraint , construct the Lagrangian function:

[0031] ;

[0032] For which 、 b 、 as well as Calculate the partial derivatives respectively and set them to zero, then bring them into the constraints of the Lagrangian function and transform the problem into solving the following linear equations:

[0033] ;

[0034] in, is a vector of all 1s; is the identity matrix; yes The kernel matrix of , is the kernel function; , , calculate the model parameters and b;

[0035] The regression function of the nadir total delay error prediction model is defined as:

[0036] ;

[0037] Radial basis function kernel Perform the solution.

[0038] As an optional implementation, the process of correcting the time-varying error of sound velocity in long-baseline underwater acoustic navigation using the trained nadir total delay error prediction model includes: using the position and velocity of the transducer at the time of signal transmission as unknown parameters, expressing the observation equation of underwater one-way acoustic navigation, applying the trained nadir total delay error prediction model to the observation equation, and obtaining a corrected observation equation.

[0039] As an optional implementation, the process of using an extended Kalman filter for navigation solution based on the corrected time-varying error of the speed of sound includes: initializing the state and error, entering the prediction phase, predicting the current state, and updating the error range, and then entering the update phase to correct the prediction results using actual measurements.

[0040] An underwater acoustic navigation enhancement system, comprising:

[0041] The joint estimation module is configured to perform standardized preprocessing on the acquired sonar data and seawater parameter data, divide the preprocessed data into blocks, each data block corresponds to a nadir total delay parameter, construct and solve the time observation equation taking into account the propagation time of the sound wave, and jointly estimate the azimuth seabed beacon position and the nadir total delay error parameter;

[0042] A nadir total delay error prediction module is configured to use the estimated nadir total delay error parameters and their corresponding time series as a training sample set, and to construct and train a nadir total delay error prediction model using a least squares support vector machine method. The nadir total delay error prediction model takes minimizing the sum of double absolute values ​​of all error variables and weight vectors as an objective function;

[0043] The correction module is configured to use the trained nadir total delay error prediction model to correct the time-varying error of sound speed in long baseline underwater acoustic navigation, and use extended Kalman filtering to perform navigation solution based on the corrected time-varying error of sound speed.

[0044] A computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, the steps in the above method are completed.

[0045] An electronic device includes a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps in the above method are completed.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] By establishing an NTD error prediction model, the present invention explicitly introduces systematic errors into the observation model and performs real-time calculations within each epoch, which is expected to more effectively compensate for the impact of sound speed changes. In addition, the present invention is based on the least squares support vector machine (LSSVM), which has strong nonlinear fitting capabilities and good generalization performance. It can achieve high-precision regression prediction under limited sample conditions without relying on complex prior models.

[0048] The present invention can significantly improve the accuracy of long-baseline underwater acoustic navigation and provide new ideas for high-precision underwater positioning in complex ocean environments.

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0051] Figure 1 The present invention is a method flow chart of an embodiment. DETAILED DESCRIPTION

[0052] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0053] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0054] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0055] Example 1

[0056] A method for enhancing underwater acoustic navigation comprises the following steps:

[0057] The acquired sonar data and seawater parameter data are standardized and preprocessed. The preprocessed data are then divided into blocks, each corresponding to a nadir total delay parameter. Taking into account the propagation time of the sound wave, a time observation equation is constructed and solved to jointly estimate the azimuth seabed beacon position and the nadir total delay error parameter.

[0058] The estimated nadir total delay error parameters and their corresponding time series are used as training sample sets, and a least squares support vector machine method is used to construct and train a nadir total delay error prediction model. The nadir total delay error prediction model takes minimizing the sum of the double absolute values ​​of all error variables and weight vectors as the objective function;

[0059] The trained nadir total delay error prediction model is used to correct the time-varying error of sound speed in long baseline underwater acoustic navigation, and an extended Kalman filter is used to perform navigation solution based on the corrected time-varying error of sound speed.

[0060] The following describes each step in detail.

[0061] The first is the joint estimation of NTD based on the GNSS-sonar (i.e., GNSS-A) system.

[0062] The position of the seafloor transponder is ultimately estimated by minimizing the residual between the observed time and the simulated propagation time. GNSS-A observations typically use a reflection transponder system to measure the round-trip propagation time of the acoustic signal, effectively eliminating the influence of external factors such as ocean current disturbances and clock drift.

[0063] (1) Data preprocessing and segmentation

[0064] After removing outliers, to improve computational stability and reduce the impact of data noise, this embodiment groups the calibration data into independent blocks of 100 data points. Each block corresponds to a specific NTD parameter, and it is assumed that the NTD remains constant over a set time period. In other words, within this short time window, the time-varying characteristics of the speed of sound can be approximately considered uniform. This block-based approach not only reduces computational complexity but also enhances the robustness of the NTD estimation, reducing interference from transient fluctuations.

[0065] (2) Constructing the observation equation

[0066] Considering the acoustic ranging data The first transponder The propagation time of the number of transmissions, the time observation equation can be written as: (1)

[0067] In the above formula, and are the observed and calculated round trip times, respectively. Indicates the i Subsea transponder locations, represents the position of the shipborne transducer, represents the lever arm offset, and Represent the observation time when the transducer transmits and receives respectively. Reference sound velocity profile Obtained by sound velocity measuring instrument. is the time delay caused by the assumption of horizontal stratification of seawater.

[0068] The propagation time delay is described by defining the gradient, which is the inverse of the speed of sound. Considering the slowness perturbation of the reference speed of sound profile, the propagation time delay is expressed as:

[0069] (2)

[0070] In the above formula, is the round-trip acoustic path, is the derivative along the acoustic path. Under time-varying horizontal stratification, the perturbation component of the slowness Depends on the depth z and time t, i.e. For simplicity, the sound path is approximated as a straight line, and the derivative along the sound wave path is expressed as Use angle of incidence According to this approximation, with time variation The propagation time delay of is written as:

[0071] (3)

[0072] In the above formula, is the travel time delay integrated along the lowest point of the sea platform, corresponding to the defined NTD parameter. After using the NTD parameter, Equation (1) can be rewritten as:

[0073] (4)

[0074] (5)

[0075] In the above formula, Corresponding to ,coefficient is a scaling factor. It should be noted that in order to accurately estimate NTD, the propagation times from multiple transponders need to be collected.

[0076] (3) Joint solution and iteration

[0077] For all observations in a data block, the following abbreviated residual equation can be constructed:

[0078] (6)

[0079] Arrange the equations for all observation data into a linear system:

[0080] (7)

[0081] In the above formula, is the unknown parameter vector, which contains the seabed benchmark coordinate correction and the NTD parameters of each data block; is the design matrix; is the observation vector.

[0082] Finally, the weighted least squares method is used to solve:

[0083] (8)

[0084] In the above formula, is the weight matrix, which can be the identity matrix initially.

[0085] Update through iteration , until If it is less than a predetermined threshold, the NTD error estimation value can be obtained.

[0086] NTD prediction based on LSSVM

[0087] To predict the temporal evolution of NTD, a method based on LSSVM was employed. This method transforms the complexity of model construction into a linear least squares problem in a high-dimensional space. By selecting a kernel function and training samples, a function that reflects the temporal evolution of NTD was constructed. LSSVM reframes the original minimization problem to optimize the model parameters.

[0088] Assume there is training samples ,in, Represents the input training sample data, Represents model output data.

[0089] (1) Optimizing the original problem

[0090] The LSSVM model represented in the feature space is as follows:

[0091] (9)

[0092] Based on formula (9), the objective function is constructed as:

[0093] (10)

[0094] Subject to equality constraints:

[0095] (11)

[0096] In the above formula, is the weight vector, Is a mapping function, which represents the mapping function from input space to feature space; represents the model bias parameter, represents the error variable; is the regularization parameter.

[0097] (2) Constructing Lagrangian function

[0098] In order to solve the above constrained optimization problem, by combining Equation (10) and Equation (11), the Lagrange factor is introduced into each constraint. , construct the Lagrangian function:

[0099] (12)

[0100] In the above formula, the interpretation of each parameter is basically the same as that of formula (10) and formula (11), and will not be elaborated here.

[0101] (3) Solving the optimal conditions

[0102] For formula (12) 、 、 as well as Find the partial derivatives separately and set them to zero, and we get the following conditions:

[0103] (13)

[0104] Substituting Equation (13) into the constraints of Equation (12), we can eliminate and , and use the defined kernel function , and finally the problem can be transformed into solving the following linear equations:

[0105] (14)

[0106] In the above formula, is a vector of all 1s; is the identity matrix; yes The kernel matrix of ; , .

[0107] After solving the above linear equations (14), the model parameters can be obtained and According to Mercer theory, for new input samples, the regression function of LSSVM is defined as:

[0108] (15)

[0109] The Radial Basis Function (RBF Kernel) is used for solving, which has an infinite-dimensional feature space and is very suitable for dealing with complex nonlinear problems. Generally defined as:

[0110] (16)

[0111] In the above formula, and t are variables; parameters is the scale parameter of the kernel function.

[0112] To identify the intrinsic relationship between NTD error and observation time, we used the GNSS-A calibration model to jointly estimate the NTD error and the position of seafloor beacons in the four directions of north, south, west, and east. The estimated NTD parameters and the corresponding Coordinated Universal Time (UTC) were used as a training dataset. The LSSVM algorithm described above was used to train this dataset and establish an NTD error prediction model.

[0113] Then enhance the LBL underwater acoustic navigation algorithm.

[0114] Using synchronous beacon mode, the position of the transducer when the signal is transmitted and speed As unknown parameters, the observation model of underwater one-way acoustic navigation can be expressed as:

[0115] (17)

[0116] Where, represents the one-way distance observation between the transducer and the i-th seafloor transponder at the k-th observation epoch; represents the Euclidean distance between the transducer and the transponder, where and represent the positions of the transducer when transmitting and receiving the acoustic signal at the kth observation epoch, respectively; represents the position of the i-th transponder on the seabed; and are systematic error and random error respectively. and denote the propagation time and sound velocity values ​​of the i-th seafloor transponder in the k-th observation epoch, respectively; represents the carrier velocity at the kth observation epoch; Indicates the time from the time the transponder transmits the signal to the time the transducer receives the signal in the kth observation epoch. i The time difference between the first and second transponder signals.

[0117] The constant velocity model dynamics model is selected to construct the state space equation of the underwater vehicle:

[0118] (18)

[0119] (19)

[0120] (20)

[0121] (twenty one)

[0122] (twenty two)

[0123] Where, represents the state transition matrix of the k-th observation epoch; represents the time interval between the kth and k-1th transmission signals of the transducer; represents the state of the carrier at the time of signal emission at the k-1th observation epoch, where 、 Respectively represent the position and speed of the carrier at that moment; represents the predicted state of the carrier at the time of signal emission at the kth observation epoch, and They represent the predicted position and predicted speed of the underwater vehicle respectively; represents the noise vector.

[0124] Applying the trained NTD prediction model to the observation equation (17), the corrected observation equation is expressed as:

[0125] (twenty three)

[0126] (twenty four)

[0127] (25)

[0128] Where, represents the time interval from the moment the seabed beacon transmits the acoustic signal in the kth observation epoch to the moment the underwater vehicle receives the i-th beacon signal; Represents the mapping function of NTD, coefficient is a scaling factor determined by the angle of incidence of the acoustic ray path Sure, Indicates a specific constant; is the sound velocity value of the i-th seabed transponder at the k-th observation moment.

[0129] By linearizing Equation (24), the observation model of the kth epoch can be obtained as follows:

[0130] (26)

[0131] (27)

[0132] (28)

[0133] Where, represents the measurement vector of the kth observation epoch, represents the Jacobian matrix of the k-th observation epoch, represents the measurement error vector of the k-th observation epoch, n represents the number of beacon signals received by the underwater vehicle in the k-th epoch, and the other symbols are consistent with the previous formula.

[0134] Next, the Kalman filter method can be used for data processing. First, the system state and error are initialized. Then, the prediction phase begins, using a constant velocity kinematic model to predict the current state and update the error range. The update phase then uses actual measurements to correct the predictions and improve estimation accuracy. This entire process is iterative, combining prediction and measurement information to effectively suppress noise interference and achieve continuous and stable estimation of the underwater vehicle state.

[0135] Example 2

[0136] An underwater acoustic navigation enhancement system, comprising:

[0137] The joint estimation module is configured to perform standardized preprocessing on the acquired sonar data and seawater parameter data, divide the preprocessed data into blocks, each data block corresponds to a nadir total delay parameter, construct and solve the time observation equation taking into account the propagation time of the sound wave, and jointly estimate the azimuth seabed beacon position and the nadir total delay error parameter;

[0138] A nadir total delay error prediction module is configured to use the estimated nadir total delay error parameters and their corresponding time series as a training sample set, and to construct and train a nadir total delay error prediction model using a least squares support vector machine method. The nadir total delay error prediction model takes minimizing the sum of double absolute values ​​of all error variables and weight vectors as an objective function;

[0139] The correction module is configured to use the trained nadir total delay error prediction model to correct the time-varying error of sound speed in long baseline underwater acoustic navigation, and use extended Kalman filtering to perform navigation solution based on the corrected time-varying error of sound speed.

[0140] Example 3

[0141] A computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, the steps in the above method are completed.

[0142] Example 4

[0143] An electronic device includes a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps in the above method are completed.

[0144] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of one or more computer-usable storage media (including but not limited to disk storage, CD - ROM , optical storage, etc.).

[0145] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0146] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0147] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0148] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made by those skilled in the art that fall within the spirit and principles of the present invention and do not require creative effort are intended to be within the scope of protection of the present invention.

Claims

1. A method for enhancing underwater acoustic navigation, characterized in that: The following steps are involved: The acquired sonar data and seawater parameter data are standardized and preprocessed. The preprocessed data are then divided into blocks, each corresponding to a nadir total delay parameter. Taking into account the propagation time of the sound wave, a time observation equation is constructed and solved to jointly estimate the azimuth seabed beacon position and the nadir total delay error parameter. The estimated nadir total delay error parameters and their corresponding time series are used as training sample sets, and a least squares support vector machine method is used to construct and train a nadir total delay error prediction model. The nadir total delay error prediction model takes minimizing the sum of the double absolute values ​​of all error variables and weight vectors as the objective function; The trained nadir total delay error prediction model is used to correct the time-varying sound velocity error in long-baseline underwater acoustic navigation, and an extended Kalman filter is used to perform navigation solution based on the corrected time-varying sound velocity error; The process of constructing the nadir total delay error prediction model using the least squares support vector machine method includes: constructing the least squares support vector machine model based on the feature space representation and constructing the objective function as follows: ； The above objective function is subject to the equality constraints: ； in, Represents the model output data, is the weight vector, Is a mapping function, which represents the mapping function from input space to feature space; b represents the model bias parameter, represents the error variable; is the regularization parameter; Then introduce the Lagrangian factor to the constraint , construct the Lagrangian function: ； For which 、 b 、 as well as Calculate the partial derivatives respectively and set them to zero, then bring them into the constraints of the Lagrangian function and transform the problem into solving the following linear equations: ； in, is a vector of all 1s; is the identity matrix; yes The kernel matrix of , is the kernel function; , , calculate the model parameters and b; The regression function of the nadir total delay error prediction model is defined as: ； Radial basis function kernel Perform the solution.

2. The underwater acoustic navigation enhancement method according to claim 1, wherein: The process of segmenting the preprocessed data into blocks includes: dividing each set number of data points into an independent data block, each data block corresponds to a nadir total delay parameter, and assuming that the nadir total delay parameter remains constant within a set time period, that is, within this time period, the time-varying characteristics of the sound speed are considered to be uniformly distributed.

3. The underwater acoustic navigation enhancement method according to claim 1, wherein: The process of constructing the time observation equation includes: taking into account the first i The propagation time of the kth transmission of a transponder is expressed as the time observation equation using the nadir total delay NTD parameter: ； ； in, and are the observed and calculated round trip times, respectively, represents the position of the i-th seabed transponder, represents the position of the shipborne transducer, represents the lever arm offset, and Respectively represent the observation time of the shipborne transducer when transmitting and receiving, and the reference sound speed profile Obtained by sound velocity measuring instrument, is the travel time delay integrated along the lowest point of the platform, corresponding to the defined nadir delay parameter, Corresponding to ,coefficient is a scaling factor determined by the angle of incidence of the acoustic ray path Sure.

4. The underwater acoustic navigation enhancement method according to claim 3, wherein: The solution process of the time observation equation includes: for all observation data in a data block, construct a simplified residual equation: ； Arrange the equations for all observation data into a linear system: ； In the above formula, X is the unknown parameter vector, which includes the seabed reference point coordinate correction and the nadir total delay NTD parameter of each data block; A is the design matrix; LS is the observation vector; Solve using weighted least squares method: ； In the above formula, P Is the weight matrix, which can be the unit matrix initially; Update through iteration X , until X If the maximum absolute value of is less than a predetermined threshold, the NTD error estimation value can be obtained.

5. The underwater acoustic navigation enhancement method according to claim 1, wherein: The process of correcting the time-varying error of sound velocity in long-baseline underwater acoustic navigation using the trained nadir total delay error prediction model includes: taking the position and velocity of the transducer at the time of signal transmission as unknown parameters, expressing the observation equation of underwater one-way acoustic navigation, applying the trained nadir total delay error prediction model to the observation equation, and obtaining a corrected observation equation.

6. The underwater acoustic navigation enhancement method according to claim 1, wherein: Based on the corrected time-varying error of the speed of sound, the process of using the extended Kalman filter for navigation solution includes: initializing the state and error, entering the prediction phase, predicting the current state and updating the error range, and then entering the update phase to correct the prediction results using the actual measurement values.

7. An underwater acoustic navigation enhancement system, applying the underwater acoustic navigation enhancement method according to claim 1, characterized in that: include: The joint estimation module is configured to perform standardized preprocessing on the acquired sonar data and seawater parameter data, divide the preprocessed data into blocks, each data block corresponds to a nadir total delay parameter, construct and solve the time observation equation taking into account the propagation time of the sound wave, and jointly estimate the azimuth seabed beacon position and the nadir total delay error parameter; A nadir total delay error prediction module is configured to use the estimated nadir total delay error parameters and their corresponding time series as a training sample set, and to construct and train a nadir total delay error prediction model using a least squares support vector machine method. The nadir total delay error prediction model takes minimizing the sum of double absolute values ​​of all error variables and weight vectors as an objective function; The correction module is configured to use the trained nadir total delay error prediction model to correct the time-varying error of sound speed in long baseline underwater acoustic navigation, and use extended Kalman filtering to perform navigation solution based on the corrected time-varying error of sound speed.

8. A computer-readable storage medium, characterized in that: Used to store computer instructions, which, when executed by a processor, complete the steps of the method according to any one of claims 1 to 6.

9. An electronic device, characterized in that: The method comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein the steps of the method according to any one of claims 1 to 6 are completed when the computer instructions are executed by the processor.

Images