Time-varying sound velocity error correction method and system based on least square support vector machine

Abstract

The invention belongs to the field of underwater acoustic positioning, and provides a time-varying sound velocity error correction method and system based on a least square support vector machine, and the method comprises the steps: constructing a sample data set through a local actual measurement sound velocity profile and corresponding actual measurement time, and dividing the sample data set to obtain a training set and a verification set; training a least square support vector machine model by using the training set, and verifying by using the verification set to obtain a trained least square support vector machine model; determining a plurality of target prediction times based on the calibrated prediction time period, and predicting a sound velocity profile for each target prediction time by using the trained least square support vector machine model; and correcting the sound velocity time-varying error at the corresponding moment in the prediction time period by using the predicted sound velocity profile of each target prediction time to obtain the corrected actually measured sound velocity at the corresponding moment in the prediction time period. The method can significantly correct the sound velocity time-varying error, thereby improving the calibration precision of the seabed reference point.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of underwater acoustic positioning, and particularly relates to a time-varying sound speed error correction method and system based on a least squares support vector machine. BACKGROUND

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

[0003] High-precision underwater acoustic positioning requires accurate sound speed profile data. Generally, sound speed profiles (SSP) can be obtained by measuring with a CTD or a sound speed profiler. However, it is difficult to cover a certain sea area to obtain continuous spatial and temporal sound speed profile changes in actual operations. If a substitute sound speed profile is used, there will be a representative error of the sound speed profile. In view of this, Zhou Shihong et al. proposed an innovative method to characterize the complexity of the ocean sound speed field by using less than 5 empirical orthogonal function analysis (EOF), which achieved an accuracy of more than 90% in expressing the sound speed field within a 0.5°x0.5° latitude and longitude grid. Based on this theoretical basis, they successfully constructed a mathematical model to describe the sound speed profile in a marine area. Ai Feng used the adaptive and self-learning ability of the backpropagation (BP) neural network and used historical sound speed data as training samples to successfully train a sound speed profile inversion prediction model. This model can use the deep learning ability of the neural network to extract features from complex historical data, and then achieve high-precision prediction of the ocean sound speed profile. Compared with using an average sound speed profile to represent the on-site results, the sound speed profile obtained by this inversion method is closer to the true value, and the main sound speed difference is concentrated in the ocean surface layer and the temperature jump layer. In addition, Wei Huang et al. proposed an empirical-orthogonal-function-based partial sound speed profile matching extension (EOF-PSSP-ME) method, which can quickly estimate the full-depth sound speed profile based on partial prior information, and can also extend the sound speed profile, with an extended average RMSE result of less than 0.1 m / s. The core idea of support vector machine (SVM) is to find an optimal hyperplane in the feature space to distinguish different classes. Later scholars proposed v-support vector machines, fuzzy support vector machines, and least squares support vector machines. The machine learning method of support vector machines provides a feasible idea for constructing a time-varying sound speed field model.

[0004] For how to suppress the sound speed error problem in underwater reference positioning, domestic and foreign scholars have carried out in-depth research on the underwater reference calibration model error correction. Ikuta et al. combined the time variation of sound speed with smoothness, used spline function to model the time delay error caused by sound speed variation, and realized the repeat observation accuracy of 10 cm in plane error and 20 cm in vertical error. Yasuda et al. analyzed the interplate locking condition based on the seafloor observation data of Japan Suruga Trough and Nankai Trough. Chen and Ikuta et al. studied the back-arc spreading of Okinawa Trough based on GNSS-acoustic observation data. Kido et al. analyzed the characteristics and rules of sound speed time variation based on GPS / acoustic and bathymetric data, and found that the amplitudes of sound speed variation of the two methods were highly consistent. Honsho and Kido learned from the GNSS troposphere delay data processing strategy, first proposed the concept of nadir total delay (NTD), and introduced cubic B-spline function to represent the system error caused by time-varying sound speed error.

[0005] Sound speed is an important factor affecting the accuracy of underwater acoustic navigation and positioning, and its representative error will cause navigation and positioning error. Although the existing research has weakened the influence of sound speed error to a certain extent, there are still various problems. The sound speed profile can be obtained by using the temperature-salinity-depth instrument, but it is difficult to cover a certain sea area to obtain continuous spatiotemporal sound speed profile variation in actual operation. Based on the existing SSP, a regional spatiotemporal sound field needs more SSPs, and the obtained SSPs have deviations from the actual ones. Using mathematical modeling to describe the sound speed variation error lacks actual data support, and the compensation accuracy of time-varying error caused by sound speed time variation is low. SUMMARY

[0006] In order to solve the above problems, the present application provides a time-varying sound speed error correction method and system based on least squares support vector machine. The application is applied to the error correction of reference calibration sound speed time-varying error. The method first trains the LSSVM model using the measured SSP, and then uses the model to predict the sound speed profile based on the current time of calibration for calculating the real-time sound speed. The method can significantly correct the sound speed time-varying error, and further improve the accuracy of seafloor reference point calibration.

[0007] According to some embodiments, the first aspect of the present application provides a time-varying sound speed error correction method based on least squares support vector machine, which adopts the following technical scheme: The time-varying sound speed error correction method based on least squares support vector machine comprises: A sample data set is constructed by using the locally measured sound speed profile and the corresponding measured time, and the sample data set is divided to obtain a training set and a validation set; training the least squares support vector machine model by using the training set and verifying by using the verification set, to obtain the trained least squares support vector machine model; determining a plurality of target prediction times based on the calibrated prediction time period, and predicting the sound velocity profile for each target prediction time by using the trained least squares support vector machine model; correcting the sound velocity time-varying error of the corresponding time of the prediction time period by using the predicted sound velocity profile of each target prediction time, to obtain the corrected measured sound velocity at the corresponding time within the prediction time period.

[0008] Further, a sample data set is constructed by using the local measured sound velocity profile and the measured time, and the sample data set is divided to obtain the training set and the verification set, specifically: obtaining the local measured sound velocity profile and the corresponding depth, and constructing a sound velocity profile data set; determining the measured time corresponding to each sound velocity profile, and constructing a time sequence; constructing a sample data set by using the sound velocity profile data set and the time sequence; dividing the sample data set according to a proportion to obtain the training set and the verification set.

[0009] Further, the least squares support vector machine model is trained by using the training set, and verified by using the verification set, to obtain the trained least squares support vector machine model, specifically: inputting the training set into the least squares support vector machine model for training to determine the corresponding relationship between the measured time and the sound velocity profile; verifying the trained least squares support vector machine model by using the verification set, and adaptively adjusting the parameters to finally obtain the trained least squares support vector machine model.

[0010] Further, the plurality of target prediction times are determined based on the calibrated prediction time period, and the sound velocity profile for each target prediction time is predicted by using the trained least squares support vector machine model, specifically: each epoch in the calibrated prediction time period is taken as a target prediction time to obtain a plurality of target prediction times; inputting each target prediction time into the trained least squares support vector machine model to output the predicted sound velocity profile corresponding to each target prediction time.

[0011] Further, the predicted sound velocity profile of each target prediction time is used to correct the sound velocity time-varying error of the corresponding time of the prediction time period, to obtain the corrected measured sound velocity at the corresponding time within the prediction time period, specifically: based on the prediction time period, the actual sound velocity is measured to obtain the depth of the transducer and the depth of the transponder corresponding to each time; The predicted sound velocity between the depth of the transducer and the depth of the transponder is extracted from the predicted sound velocity profile to obtain the extracted predicted sound velocity profile and the corresponding number of water layers and the overall depth of the water body. The equivalent average sound velocity at each moment is determined by using the intercepted predicted sound velocity profile at each moment, the corresponding number of intercepted water layers, and the overall depth of the intercepted water body. The equivalent average speed of sound at each moment is used as the corrected measured speed of sound for the prediction time period.

[0012] Furthermore, the calculation of the equivalent average sound speed is specifically as follows: ; in, For the first One-way delay per epoch; It is the first The equivalent average speed of sound in each epoch; It is the first Predicted sound velocity profile at each epoch Overall water depth value, and They are The first interception of water bodies Layers and Layer depth value, and yes The speed of sound corresponding to the water layer, It refers to the number of water layers intercepted.

[0013] According to some embodiments, a second aspect of the present invention provides a time-varying sound speed error correction system based on least squares support vector machines, employing the following technical solution: A time-varying sound speed error correction system based on least squares support vector machines includes: The dataset partitioning module is configured to construct a sample dataset using local measured sound velocity profiles and corresponding measured times, and to partition the sample dataset to obtain a training set and a validation set. The model training module is configured to train a least squares support vector machine model using the training set and validate it using the validation set to obtain a trained least squares support vector machine model. The sound velocity profile prediction module is configured to determine multiple target prediction times based on a calibrated prediction time period, and use a trained least squares support vector machine model to predict the sound velocity profile for each target prediction time. The sound velocity error correction module is configured to use the predicted sound velocity profile of each target prediction time to correct the time-varying error of the sound velocity at the corresponding moment in the prediction time period, so as to obtain the corrected measured sound velocity at the corresponding moment in the prediction time period.

[0014] According to some embodiments, a third aspect of the present application provides a computer readable storage medium.

[0015] A computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the time-varying sound speed error correction method based on least squares support vector machine according to the first aspect.

[0016] According to some embodiments, a fourth aspect of the present application provides a computer device.

[0017] A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the time-varying sound speed error correction method based on least squares support vector machine according to the first aspect when executing the program.

[0018] According to some embodiments, a fifth aspect of the present application provides a computer program product or computer program.

[0019] A computer program product or computer program comprising computer instructions stored in a computer readable storage medium, wherein a processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to cause the computer device to perform the steps of the time-varying sound speed error correction method based on least squares support vector machine according to the first aspect.

[0020] Compared with the prior art, the present application has the following beneficial effects: The present application proposes a time-varying sound speed error real-time correction method based on LSSVM in view of the limitations of the prior art and the actual situation of machine learning algorithm. The method uses local measured SSP to train the LSSVM model to predict the sound speed profile. We elaborate how to construct the data set to train the model, and then predict the sound speed profile according to the time period of the calibrated measured data, and find the nearest SSP based on the current time in the seabed benchmark calibration algorithm to apply to the sound speed calculation. This method can correct the time-varying sound speed error, thereby improving the positioning accuracy of the underwater benchmark point. BRIEF DESCRIPTION OF DRAWINGS

[0021] The drawings constituting a part of this application are used to provide a further understanding of the present application, and the illustrative embodiments of the present application and their descriptions are used to explain the present application, and do not constitute an improper limitation on the present application.

[0022] Figure 1 is a flowchart of the time-varying sound speed error correction method based on least squares support vector machine in the embodiments of the present application; Figure 2 is a flowchart of the LSSVM prediction of sound speed profile in the embodiments of the present application. Detailed Implementation

[0023] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0024] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0025] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0026] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0027] Terminology Explanation: SSP: Sound Speed ​​Profile; EOF: Empirical Orthogonal Function. LSSVM: Least Squares Support Vector Machine.

[0028] Example 1 like Figure 1 As shown, this embodiment provides a time-varying sound speed error correction method based on least squares support vector machines. This embodiment uses the application of this method to a server as an example for illustration. It is understood that this method can also be applied to terminals, and can also be applied to systems including terminals, servers, and other components, and implemented through interaction between the terminal and the server. The server can be an independent physical server, a server cluster composed of multiple physical servers, or a distributed system. It can also be a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network servers, cloud communication, middleware services, domain name services, CDN security services, and big data and artificial intelligence platforms. The terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, etc., but is not limited to these. The terminal and server can be directly or indirectly connected via wired or wireless communication, which is not limited herein. In this embodiment, the method includes the following steps: Step S1: Constructing a sample data set by using the local measured sound velocity profile and the corresponding measured time, and dividing the sample data set to obtain a training set and a validation set; Step S2: Training a least squares support vector machine model by using the training set, and verifying by using the validation set to obtain the trained least squares support vector machine model; Step S3: Determining a plurality of target prediction times based on the calibrated prediction time period, and predicting the sound velocity profile for each target prediction time by using the trained least squares support vector machine model; Step S4: Correcting the sound velocity time-varying error of the corresponding time of the prediction time period by using the predicted sound velocity profile of each target prediction time to obtain the corrected measured sound velocity at the corresponding time in the prediction time period.

[0029] Specifically, the least squares support vector machine (LSSVM) is an improved type of traditional support vector machine, which is based on statistical theory and converts a quadratic optimization problem into the solution of a linear equation set. This method constructs a function reflecting the characteristics of sound velocity change by selecting a kernel function and training samples, as follows: Let the training data set be , which satisfies the function: (1); wherein, is the coefficient vector is the error between the predicted value and the true value, is the bias term.

[0030] According to the least squares principle, the problem is optimized as: (2); is the adjustment coefficient.

[0031] The Lagrange function is constructed as follows: (3); wherein, is the Lagrange coefficient.

[0032] Take partial derivatives of , , , respectively, and set them equal to 0 to obtain: (4); The above formula can be simplified as: (5); wherein, , , , .

[0033] The regression function can be obtained as: (6); As shown in Figure 2 , the method of the embodiment specifically comprises: Step S1: using local measured sound velocity profile and measured time to construct a sample data set, and dividing the sample data set to obtain a training set and a validation set, specifically: Obtaining a local measured sound velocity profile and its corresponding depth, and constructing a sound velocity profile data set; Determining the measured time corresponding to each sound velocity profile, and constructing a time sequence; Constructing a sample data set from the sound velocity profile data set and the time sequence; Divide the sample data set according to the proportion to obtain the training set and the validation set.

[0034] Based on the principle of the least squares support vector machine LSSVM model in the above formulas (1)-(6), when using the model, the training input sample includes the measured time sequence and the corresponding sound velocity profile data set , The number of sound velocity profiles.

[0035] Wherein, the measured time corresponding sound velocity profile , specifically: (7); Wherein, , is the depth, is the total number of layers of the water body, , is the sound velocity corresponding to the depth.

[0036] Step S2: training the least squares support vector machine model using the training set, and verifying using the validation set to obtain the trained least squares support vector machine model.

[0037] Specifically, the training set is input into the least squares support vector machine model for training to determine the corresponding relationship between the measured time and the sound velocity profile, and then the validation set is used to verify the trained LSSVM model and adaptively adjust the parameters, and finally the trained least squares support vector machine model (LSSVM model) is obtained.

[0038] Step S3: determining a plurality of target prediction times based on the calibrated prediction time period, and using the trained least squares support vector machine model to predict the sound velocity profile for each target prediction time, specifically: each epoch in the calibrated prediction time period as a target prediction time, to obtain a plurality of target prediction times; input each target prediction time into the trained least squares support vector machine model, and output a predicted sound speed profile corresponding to each target prediction time.

[0039] set the calibrated prediction time period , set the time corresponding to the i-th epoch as the target prediction time , in the training set, there are a measured sound speed profile data set and a corresponding measured time sequence , then the predicted sound speed profile at this time is: (8); wherein, is a prediction function based on the trained LSSVM model, that is, based on the relationship between the measured time sequence , the measured sound speed profile data set and the target prediction time , to determine the predicted sound speed profile.

[0040] , the predicted sound speed profile corresponding to the i-th epoch is , and the specific process is as follows: (9); wherein, , is the depth, , is the predicted sound speed corresponding to the depth.

[0041] In step S4, the predicted sound speed profile at each target prediction time is used to correct the sound speed time-varying error at the corresponding time of the prediction time period, to obtain the corrected measured sound speed at the corresponding time in the prediction time period, and the specific process is as follows: Based on the prediction time period, the actual sound speed is measured to obtain the depth of the transducer and the depth of the transponder corresponding to each time; In the predicted sound speed profile, the predicted sound speed between the depth of the transducer and the depth of the transponder is intercepted to obtain the intercepted predicted sound speed profile and the corresponding intercepted water layer number and the intercepted water body overall depth value; Using the intercepted predicted sound speed profile at each time and the corresponding intercepted water layer number and the intercepted water body overall depth value, the equivalent average sound speed at each time is determined; The equivalent average sound speed at each time is used as the corrected measured sound speed in the prediction time period.

[0042] ​When the sound speed variation error is corrected by using the predicted sound speed profile, the position of the seabed reference point is determined by the way of distance intersection in underwater positioning. Generally, the position of the GNSS antenna of the waterborne carrier (surveying ship) is determined first, and then the coordinates of the shipborne transducer are converted according to the attitude parameters. The distance observation value is obtained by measuring the propagation time delay of the sound signal between the transducer and the seabed transponder and the sound speed. Finally, the coordinates of the seabed reference point are solved by the method of distance intersection.

[0043] Generally, the underwater positioning distance observation equation is: (9); The embodiment proposes to calculate the equivalent sound speed based on the predicted sound speed profile, and then to calibrate the reference point. At this time, the measured sound speed at each time in the predicted time period of calibration is replaced by the equivalent average sound speed calculated by the predicted sound speed profile of the first epoch. The calculation method of the equivalent average sound speed is as follows: The predicted sound speed profile data between the depth of the transducer and the depth of the transponder at each time in the predicted time period of calibration is intercepted, and the sound speed at the depth of the transducer and the depth of the transponder is obtained by linear interpolation of the adjacent sound speed data on the depth. The water body level of the intercepted sound speed profile data is divided into N layers according to the current time; The weighted equivalent average sound speed value is calculated by the following formula: (11); Then the underwater positioning distance observation equation should be: (12); (13); Wherein, is the one-way time delay of the first epoch; is the equivalent average sound speed of the first epoch; is the straight line distance between the shipborne transducer and the seabed transponder; is the coordinate of the shipborne transducer at the first epoch; is the coordinate of the seabed transponder; is the system error in the measurement process, including the installation and calibration deviation of the transducer, the response error of the transponder, the sound speed disturbance error, etc. is the random error. is the first layer of the water body; is the second layer of the water body. is the ​A predicted sound velocity profile of an epoch a total water depth value, and are respectively depth values of the first layer and layer of the intercepted water body, and are predicted sound velocities of the corresponding layers of the water body.

[0044] Embodiment Two The embodiment provides a time-varying sound velocity error correction system based on a least squares support vector machine, comprising: a data set division module configured to construct a sample data set by using a local measured sound velocity profile and corresponding measured time, and divide the sample data set to obtain a training set and a verification set; a model training module configured to train a least squares support vector machine model by using the training set, and verify by using the verification set to obtain a trained least squares support vector machine model; a sound velocity profile prediction module configured to determine a plurality of target prediction times based on a calibrated prediction time period, and predict a sound velocity profile for each target prediction time by using the trained least squares support vector machine model; a sound velocity error correction module configured to correct a sound velocity time-varying error at a corresponding time in the prediction time period by using a predicted sound velocity profile of each target prediction time, and obtain a corrected measured sound velocity at the corresponding time in the prediction time period.

[0045] The above modules and the examples and application scenarios realized by the corresponding steps are the same, but are not limited to the content disclosed in the above embodiment one. It should be noted that the above modules, as part of the system, can be executed in a computer system such as a group of computer executable instructions.

[0046] The description of each of the above embodiments focuses on different aspects, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0047] The proposed system can be implemented in other ways. For example, the system embodiments described above are only illustrative, for example, the division of the above modules is only a logical function division, and in actual implementation, there can be another division mode, for example, a plurality of modules can be combined or integrated into another system, or some features can be ignored or not executed.

[0048] Embodiment Three The embodiment provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to realize the steps in the time-varying sound velocity error correction method based on a least squares support vector machine as described in the above embodiment one.

[0049] Embodiment Four The embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps in the time-varying sound speed error correction method based on the least squares support vector machine according to the embodiment one.

[0050] Embodiment Five The embodiment provides a computer program product or a computer program, including computer instructions stored in a computer readable storage medium, and a processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device executes the steps in the time-varying sound speed error correction method based on the least squares support vector machine according to the embodiment one.

[0051] Those skilled in the art should understand that the embodiments of the present application can provide a method, a system or a computer program product. Therefore, the present application can take the form of a hardware embodiment, a software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer usable program code.

[0052] The present application is described with reference to flowcharts and / or block diagrams of the method, device (system) and computer program product according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams and the combination of the flows and / or blocks can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices produce a device implemented in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 means for performing the function specified by one or more blocks.

[0053] These computer program instructions can also be stored in a computer readable storage medium capable of guiding the computer or other programmable data processing devices to work in a specific way, so that the instructions stored in the computer readable storage medium produce a product including instruction means, which implements the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 means for performing the function specified by one or more blocks.

[0054] These computer program instructions can also be loaded into a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 The computer program instructions can also be loaded into a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 The computer program instructions can also be loaded into a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks.

[0055] Those skilled in the art should understand that all or part of the flowcharts described above can be implemented by a computer program instructing related hardware, and the program can be stored in a computer readable storage medium. When the program is executed, it can include the flowcharts of the above-mentioned embodiments of the method. The storage medium can be a magnetic disc, an optical disc, a read-only memory (ROM) or a random access memory (RAM), etc.

[0056] Although the specific embodiments of the present application are described above with reference to the drawings, the description is not a limitation on the scope of protection of the present application. Those skilled in the art should understand that various modifications or variations can be made to the technical solutions of the present application without creative labor, and still fall within the scope of protection of the present application.

Claims

1. A time-varying sound velocity error correction method based on least squares support vector machines, characterized by, The method comprises the steps of: constructing a sample data set by using local measured sound velocity profiles and corresponding measured times, and dividing the sample data set to obtain a training set and a verification set; training a least squares support vector machine model by using the training set, and verifying by using the verification set to obtain the trained least squares support vector machine model; determining a plurality of target prediction times based on the calibrated prediction time period, and predicting a sound velocity profile for each target prediction time by using the trained least squares support vector machine model; correcting the sound velocity time-varying error of the corresponding time of the prediction time period by using the predicted sound velocity profile of each target prediction time to obtain the corrected measured sound velocity at the corresponding time in the prediction time period.

2. The least squares support vector machines based time- varying sound speed error correction method of claim 1, wherein, A sample data set is constructed by using local measured sound velocity profiles and measured times, and the sample data set is divided to obtain a training set and a verification set, specifically: obtaining local measured sound velocity profiles and corresponding depths to construct a sound velocity profile data set; determining the measured time corresponding to each sound velocity profile to construct a time sequence; constructing a sample data set from the sound velocity profile data set and the time sequence; dividing the sample data set according to a proportion to obtain a training set and a verification set.

3. The least squares support vector machines based time varying sound speed error correction method of claim 1, wherein, The least squares support vector machine model is trained by using the training set, and the verification set is verified to obtain the trained least squares support vector machine model, specifically: inputting the training set into the least squares support vector machine model for training to determine the corresponding relationship between the measured time and the sound velocity profile; verifying the trained least squares support vector machine model by using the verification set, and adaptively adjusting the parameters to finally obtain the trained least squares support vector machine model.

4. The least squares support vector machines based time- varying sound speed error correction method of claim 1, wherein, The plurality of target prediction times are determined based on the calibrated prediction time period, and the sound velocity profile for each target prediction time is predicted by using the trained least squares support vector machine model, specifically: each epoch in the calibrated prediction time period is taken as a target prediction time to obtain a plurality of target prediction times; each target prediction time is input into the trained least squares support vector machine model to output the predicted sound velocity profile corresponding to each target prediction time.

5. The least squares support vector machines based time- varying sound speed error correction method of claim 1, wherein, The predicted sound velocity profile of each target prediction time is used to correct the sound velocity time-varying error of the corresponding time of the prediction time period to obtain the corrected measured sound velocity at the corresponding time in the prediction time period, specifically: actual sound velocity measurement is performed based on the prediction time period to obtain the depth of the transducer and the depth of the transponder corresponding to each time; the predicted sound velocity between the depth of the transducer and the depth of the transponder in the predicted sound velocity profile is intercepted to obtain the intercepted predicted sound velocity profile and the corresponding intercepted water layer number and the intercepted water overall depth value; the equivalent average sound velocity at each time is determined by using the intercepted predicted sound velocity profile at each time and the corresponding intercepted water layer number and the intercepted water overall depth value; the equivalent average sound velocity at each time is taken as the corrected measured sound velocity in the prediction time period.

6. The least squares support vector machines based time- varying sound speed error correction method of claim 5, wherein, The calculation of the equivalent average sound velocity is specifically: ； in, For the first One-way delay per epoch; It is the first The equivalent average speed of sound in each epoch; It is the first Predicted sound velocity profile at each epoch Overall water depth value, and They are The first interception of water bodies Layers and Layer depth value, and yes The speed of sound corresponding to the water layer, It refers to the number of water layers intercepted.

7. A time-varying sound speed error correction system based on least squares support vector machines, characterized in that, including: a data set division module configured to construct a sample data set by using local measured sound velocity profiles and corresponding measured times, and divide the sample data set to obtain a training set and a verification set; The model training module is configured to train the least squares support vector machine model by using the training set and validate by using the validation set to obtain the trained least squares support vector machine model; The sound speed profile prediction module is configured to determine a plurality of target prediction times based on the calibrated prediction time period, and predict the sound speed profile for each target prediction time by using the trained least squares support vector machine model; The sound speed error correction module is configured to correct the sound speed time-varying error at the corresponding time in the prediction time period by using the predicted sound speed profile at each target prediction time to obtain the corrected measured sound speed at the corresponding time in the prediction time period.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to implement the steps in the least squares support vector machine-based time-varying sound speed error correction method of any one of claims 1-6.

9. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the steps in the least squares support vector machine-based time-varying sound speed error correction method of any one of claims 1-6.

10. A computer program product, characterised in that, The computer program product comprises a computer program, which, when executed by a processor, implements the steps in the least squares support vector machine-based time-varying sound speed error correction method of any one of claims 1-6.

Images