CN113312830B - Method for obtaining marine mesoscale eddy wave impedance based on deep learning and processing terminal - Google Patents
Method for obtaining marine mesoscale eddy wave impedance based on deep learning and processing terminal Download PDFInfo
- Publication number
- CN113312830B CN113312830B CN202110078654.6A CN202110078654A CN113312830B CN 113312830 B CN113312830 B CN 113312830B CN 202110078654 A CN202110078654 A CN 202110078654A CN 113312830 B CN113312830 B CN 113312830B
- Authority
- CN
- China
- Prior art keywords
- model
- mesoscale
- vortex
- deep learning
- mesoscale vortex
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000012545 processing Methods 0.000 title claims abstract description 25
- 238000013135 deep learning Methods 0.000 title claims abstract description 21
- 238000003384 imaging method Methods 0.000 claims abstract description 27
- 238000013136 deep learning model Methods 0.000 claims abstract description 24
- 239000013535 sea water Substances 0.000 claims description 34
- 238000013519 translation Methods 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 7
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 5
- 238000004088 simulation Methods 0.000 claims description 2
- 238000011835 investigation Methods 0.000 description 8
- 238000011160 research Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- NMJORVOYSJLJGU-UHFFFAOYSA-N methane clathrate Chemical compound C.C.C.C.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O NMJORVOYSJLJGU-UHFFFAOYSA-N 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000012958 reprocessing Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/08—Fluids
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses a method for obtaining marine mesoscale vortex impedance based on deep learning and a processing terminal, wherein the method comprises the steps of obtaining a plurality of mesoscale vortex impedance models through a sound velocity model and a density model; performing forward modeling on each mesoscale vortex model to obtain seismic reflection data, and then performing imaging processing to obtain a seismic reflection imaging section; constructing a seismic reflection imaging section and a mesoscale vortex wave impedance model as tag data; constructing a deep learning model, and inputting label data to obtain the trained deep learning model; and processing the actually acquired seismic data to obtain a seismic reflection imaging section to be extracted, inputting the seismic reflection imaging section to the trained deep learning model for processing, and obtaining the wave impedance of the mesoscale vortex. The method based on deep learning can extract the wave impedance information of the vortex from the weak seismic reflection signal, and the method based on deep learning can be suitable for vortex complex structures.
Description
Technical Field
The invention relates to the technical field of extracting mesoscale eddy parameters in physical oceans, in particular to a method for obtaining the mesoscale eddy wave impedance in oceans based on deep learning and a processing terminal.
Background
Mesoscale vortexes (also called mesoscale vortexes or weather type ocean vortexes) have important influence and significance in the aspects of ocean dynamics, ocean biology and ocean environment information guarantee, and are research hotspots of oceanologists at home and abroad. Currently, research on mesoscale vortexes mainly focuses on two aspects, namely, the existence of the mesoscale vortexes is determined through remote sensing and hydrological investigation, and the vortexes are identified and classified by using sea level height abnormal data, which can be generally classified into cyclone, anti-cyclone, non-cyclone and the like, so as to improve the efficiency of vortex identification and classification.
However, it is rarely related to extracting relevant basic parameters from the mesoscale vortexes, wherein the basic parameters of the mesoscale vortexes include temperature (temperature), salinity (salinity), density (density), sonic velocity field information and the like, and the basic parameters have important significance for understanding the vortexes. Conventionally, if the basic parameters need to be obtained, the basic parameters are mainly obtained through marine hydrological actual investigation, but the traditional physical marine observation is difficult to carry out wide-range observation in a short time, and further extraction of the mesoscale vortex parameters is limited. Therefore, it is necessary to obtain these basic parameters by using new scientific techniques.
The newly developed seismic oceanographic methods provide a new perspective for vortex studies. The seismic oceanography method is a method for researching oceans by using a reflection seismic method. The reflection seismic method can image the internal structure of the seawater, and a reflection seismic section after the reflection seismic data processing shows a plurality of reflection homophase axes aiming at a wave impedance interface formed by the seawater in the mesoscale vortex due to the temperature and salinity difference, so that the internal structure of the vortex can be more easily identified in detail. Meanwhile, the data interval of the reflection seismic data is very dense, and the resolution ratio is higher (the horizontal sampling interval can reach 6.25 meters). A common practice for studying the ocean using seismic methods is to extract the wave impedance parameters of the aquifer from the seismic data. The wave impedance is the product of the density and the propagation sound velocity of seismic waves, and the sound velocity in the sea water layer has good corresponding relation with temperature, pressure and the like. Once the wave impedance information is obtained, the density, temperature, salinity and the like of the vortex can be obtained through conversion of a certain formula. Thus, the wave impedance of vortices can be extracted based on reflection seismology methods to obtain parameters of mesoscale vortices.
There are two main problems associated with extracting vortex impedance parameters from seismic data: (1) due to the fact that the difference of wave impedance in the water layer is small, the seismic reflection signals are weak; (2) the boundaries and internal structure of the vortex are complex. These characteristics make it difficult for traditional seismic wave impedance inversion methods to achieve good results. To better extract wave impedance parameters from seismic data, some have proposed constraining the wave impedance inversion based on measured XBT data. A large amount of seismic data are accumulated in the edge sea and the ocean of the world, and the current and local sea water layer structure and physical parameters can be obtained by reprocessing and inverting. The XBT data corresponding to seismic lines is very small relative to this large volume of seismic data, and it is very expensive to acquire XBT data.
Disclosure of Invention
Aiming at the defects of the prior art, one of the purposes of the invention is to provide a method for obtaining the marine mesoscale eddy wave impedance based on deep learning, which can solve the problem of extracting the mesoscale eddy wave impedance parameters;
it is a second object of the present invention to provide a processing terminal that can solve the problem of extracting the mid-scale eddy wave impedance parameters.
The technical scheme for realizing one purpose of the invention is as follows: a method for obtaining marine mesoscale eddy wave impedance based on deep learning comprises the following steps:
step 1: establishing a sea water layer background sound velocity model and a sea water layer background density model, correspondingly adding a sound velocity model of a mesoscale vortex and a density model of a mesoscale vortex on the sea water layer background sound velocity model and the sea water layer background density model respectively to obtain a mesoscale vortex sound velocity model and a mesoscale vortex density model respectively,
the sound velocity model of the mesoscale vortex and the density model of the mesoscale vortex are obtained through presetting;
step 2: respectively multiplying a plurality of different mesoscale vortex sound velocity models and a plurality of different mesoscale vortex density models to obtain a plurality of mesoscale vortex wave impedance models;
and step 3: performing forward modeling on each mesoscale vortex model to obtain seismic reflection data, and performing imaging processing on all the seismic reflection data respectively to obtain corresponding seismic reflection imaging sections;
and 4, step 4: constructing a seismic reflection imaging section and a mesoscale vortex wave impedance model into tag data, wherein the tag data comprise target data and input data, the target data are the mesoscale vortex wave impedance model, and the input data are the seismic reflection imaging section;
and 5: building a deep learning model based on a neural network, dividing the label data in the step 4 into at least two parts, wherein one part is used for training the deep learning model, the other part is used for testing the deep learning model, and obtaining the finally trained deep learning model after the training and testing are qualified;
step 6: the actually collected seismic data is processed in the same way as the step 3 to obtain a seismic reflection imaging section of the wave impedance to be extracted,
and inputting the seismic reflection imaging section of the wave impedance to be extracted into the trained deep learning model for processing, wherein the result output by the deep learning model is the wave impedance of the mesoscale vortex, and the processing is finished.
Further, in step 2, the multiple different mesoscale vortex sound velocity models and the multiple different mesoscale vortex density models are obtained, and the specific implementation process is as follows:
and (2) rotating or translating the mesoscale vortex sound velocity model and/or the mesoscale vortex density model in the step (1), wherein the rotation angle is random, the translation distance is random, and a mesoscale vortex sound velocity model or a mesoscale vortex density model is correspondingly obtained in each rotation or translation, so that a plurality of different mesoscale vortex sound velocity models and a plurality of different mesoscale vortex density models are obtained.
Further, in step 3, in the forward modeling, rake wavelets with different frequencies are randomly selected as seismic sources.
Further, in the forward modeling, a forward result is obtained by solving a sound wave equation of the mesoscale vortex model through numerical values, the forward result is seismic reflection data, and the sound wave equation is as follows:
where v represents the speed of sound, P represents the density, P represents the seismic wavefield,representing the dirac operator, F the source force, and t the time.
Furthermore, the neural network is obtained by combining a convolutional layer and a full-link layer, and in the neural network, the number of nodes of an input layer and the number of nodes of an output layer are both Nz*Nx,NzNumber of sampling points in depth direction, NxIndicating the number of sample points in the horizontal lateral direction.
The second technical scheme for realizing the aim of the invention is as follows: a processing terminal, comprising:
a memory for storing program instructions;
a processor for executing the program instructions to perform the steps of the method for obtaining the mid-sea mesoscale eddy wave impedance based on deep learning.
The invention has the beneficial effects that: the method based on deep learning can extract the wave impedance information of the vortex from the weak seismic reflection signal, and the method based on deep learning can be suitable for vortex complex structures.
Drawings
FIG. 1 is an exemplary model of the background acoustic velocity of a sea water layer;
FIG. 2 is an exemplary model of background density of a sea water layer;
FIG. 3 is a sound velocity model of a mesoscale vortex of an example;
FIG. 4 is a model of the acoustic velocity of the mesoscale vortices generated after a certain degree of rotation based on FIG. 3;
FIG. 5 is a model of the acoustic velocity of the mesoscale vortices obtained after a translation of a certain distance on the basis of FIG. 3;
figure 6 is a schematic representation of a seismic reflection profile of a real mesoscale vortex in a sea area of the pacific ocean,
FIG. 7 is a graphical representation of the results obtained from the processing of FIG. 6 according to the present method;
FIG. 8 is a flow chart of the preferred embodiment;
fig. 9 is a schematic diagram of a processing terminal.
Detailed Description
The invention is further described with reference to the accompanying drawings and the specific embodiments.
As shown in fig. 1 to 8, a method for obtaining marine mesoscale eddy-wave impedance based on deep learning includes the following steps:
step 1: and establishing a sea water layer background sound velocity model and a sea water layer background density model. The seawater layer background models are established, and a seawater layer background sound velocity model and a seawater layer background density model can be preset by using the data and basic knowledge based on the actual physical marine investigation, hydrological investigation and related professional theoretical knowledge. In the prior art, much knowledge is accumulated about a sea water layer background sound velocity model and a sea water layer background density model, so that the existing data can be fully utilized. FIG. 1 is an exemplary model of the background acoustic velocity of a sea water layer, which is formulated as follows:
similarly, FIG. 2 is an exemplary model of the background density of the sea water layer, which is formulated as follows:
ρ(x,z)=1000+0.002*z
of course, the sea water layer background sound velocity model and the sea water layer background density model represented by other formulas can also be established according to other sea water parameter formulas or actually measured CTD (thermohalimeter) parameters. It should be noted that, in a real marine environment, the sound velocity parameter and the density parameter have a certain correlation, and therefore, the correlation between the sound velocity and the density needs to be considered in the established sea water layer background sound velocity model and the sea water layer background density model.
Because the wave impedance parameter is the product of the sound velocity parameter and the density parameter, the wave impedance model can be obtained only by multiplying the sea water layer background sound velocity model and the sea water layer background density model.
Step 2: and correspondingly adding a sound velocity model of the mesoscale vortex and a density model of the mesoscale vortex on the sea water layer background sound velocity model and the sea water layer background density model respectively so as to obtain the mesoscale vortex sound velocity model and the mesoscale vortex density model respectively. That is, the sound velocity model of the mesoscale vortex is added to the sea water layer background sound velocity model, and the density model of the mesoscale vortex is added to the sea water layer background density model, wherein the adding means adding of the two.
And multiplying the obtained mesoscale vortex sound velocity model and the mesoscale vortex density model to obtain a mesoscale vortex wave impedance model.
The method comprises the steps of obtaining a sound velocity model of a mesoscale vortex and a density model of the mesoscale vortex based on satellite remote sensing, field hydrological investigation and numerical mode comprehensive presetting, accumulating a large amount of knowledge about the density and the sound velocity of the mesoscale vortex in the prior art, and based on the research of seismic oceanography and the research of physical oceanography, the shape of the mesoscale vortex in the vertical direction is close to an ellipse, the scale in the transverse direction is different from dozens of kilometers, and the scale in the vertical direction is different from hundreds of meters to thousands of meters. Based on the knowledge, sound velocity models of different sizes of mesoscale vortices and density models of the mesoscale vortices can be established. FIG. 3 is a sound velocity model of a mesoscale vortex of an example.
And step 3: because each mesoscale vortex sound velocity model and the mesoscale vortex density model are multiplied to obtain a corresponding mesoscale vortex wave impedance model, a plurality of mesoscale vortex wave impedance models can be obtained when a plurality of mesoscale vortex sound velocity models and a plurality of mesoscale vortex density models exist. For example, the mesoscale vortex sound velocity model and/or the mesoscale vortex density model obtained in the step 2 are/is subjected to transformation operations such as rotation and translation, the rotation angle is random, the translation distance can also be random, and a mesoscale vortex sound velocity model or a mesoscale vortex density model is obtained correspondingly every time of rotation or translation, so that more mesoscale vortex sound velocity models and mesoscale vortex density models with different parameter values are obtained, and more mesoscale vortex wave impedance models are obtained. Similarly, specific values of the two parameters of the sound velocity and/or the density can be randomly assigned, namely, a certain value is randomly increased or decreased on the original parameters of the density and the sound velocity, so that more mesoscale vortex wave impedance models are obtained.
Fig. 4 is a sound velocity model of a mesoscale vortex obtained by rotating a certain angle on the basis of fig. 3, and fig. 5 is a sound velocity model of a mesoscale vortex obtained by translating a certain distance on the basis of fig. 3.
Through the steps, a plurality of mesoscale vortex models can be obtained, each mesoscale vortex model comprises a mesoscale vortex sound velocity model, a mesoscale vortex density model and a mesoscale vortex wave impedance model, and the mesoscale vortex model is a model comprising three groups of parameters. Where the speed of sound is denoted by v and the density by ρ.
And 4, step 4: and (4) performing forward simulation on each mesoscale vortex model obtained in the step (3) to obtain corresponding forward data, wherein the forward data is the seismic reflection data. In forward modeling, Rake wavelets with different frequencies can be randomly selected as seismic sources. The forward modeling of each mesoscale vortex model can be obtained by numerically solving an acoustic wave equation, and the solution result is forward data. The acoustic wave equation is shown in formula (r):
in the formula, P represents the seismic wavefield,representing the dirac operator, F the source force, and t the time.
And 5: and (4) respectively carrying out imaging processing on all the seismic reflection data obtained in the step (4) to obtain corresponding seismic reflection imaging sections. The imaging processing of the seismic reflection data mainly includes observation system definition, velocity analysis, pre-stack time migration, and the like, which are all in the prior art and are not described herein again.
Step 6: record P for the seismic reflection imaging section obtained from the ith mesoscale vortex modeliI.e. PiRepresenting the ith seismic reflection imaging section. Constructing a seismic reflection imaging section and a mesoscale vortex wave impedance model into tag data, wherein the tag data comprises target data and input data, the target data is the mesoscale vortex wave impedance model,the input data is a seismic reflection imaging section. The label data is also the label data set for subsequent deep learning. Wherein the seismic reflection imaging data and the vortex wave impedance model are both Nz*NxIn a two-dimensional array of (1), wherein NzNumber of sampling points in depth direction, NxThe number of sampling points in the horizontal transverse direction.
And 7: constructing a deep learning model, wherein the deep learning model adopts a neural network model combined by a convolution layer and a full connection layer, and the number of nodes of the input layer and the output layer is Nz*Nx. And training the constructed deep learning model to obtain the trained deep learning model. And dividing the label data obtained in the step 6 into two parts, wherein one part is used for training the deep learning model, the other part is used for testing the deep learning model, and the finally trained deep learning model is obtained after the training and testing are qualified.
And 8: and 5, processing the actually acquired seismic data in the same way as the step 5 to obtain a seismic reflection imaging section of the wave impedance to be extracted, inputting the seismic reflection imaging section of the wave impedance to be extracted into the trained deep learning model for processing, wherein the result output by the deep learning model is the wave impedance of the medium-scale vortex, and thus the wave impedance parameters are extracted. And converting the wave impedance information by the conventional formula to obtain basic parameter information such as density, sound velocity, temperature and the like of the mesoscale vortex.
FIG. 6 is a schematic diagram of a seismic reflection section of a real mesoscale vortex in a certain sea area of the Pacific ocean, and the result obtained by processing the seismic reflection section according to the method is shown in FIG. 7, namely, a schematic diagram of the wave impedance obtained by extraction.
The method based on deep learning can extract the wave impedance information of the vortex from the weak seismic reflection signal, and the method based on deep learning can be suitable for vortex complex structures. The method for obtaining the vortex wave impedance parameters has important significance for the research of marine dynamics, marine biology and marine environment. Also has important significance for offshore operation. The invention can be well applied to ocean engineering platform equipment, such as engineering investigation ships, ocean investigation ships, seabed resource investigation ships, hydrological measurement ships and other operation ships, so that the basic parameters of the mesoscale vortex can be well and quickly acquired, and the operation efficiency and the operation safety are improved. Particularly, in the early-stage mining preparation process of deep sea resources such as natural gas hydrate and the like, the basic parameters of the mesoscale vortex of the area to be mined are obtained to evaluate the starting safety, and the corresponding mining scheme is improved based on the obtained basic parameters of the mesoscale vortex, so that the operation efficiency is improved.
As shown in fig. 9, the invention also relates to a processing terminal 100 comprising:
a memory 101 for storing program instructions;
a processor 102 for executing the program instructions to perform the steps of the method for obtaining the mid-sea mesoscale eddy wave impedance based on deep learning.
The embodiments disclosed in this description are only an exemplification of the single-sided characteristics of the invention, and the scope of protection of the invention is not limited to these embodiments, and any other functionally equivalent embodiments fall within the scope of protection of the invention. Various other changes and modifications to the above-described embodiments and concepts will become apparent to those skilled in the art from the above description, and all such changes and modifications are intended to be included within the scope of the present invention as defined in the appended claims.
Claims (5)
1. A method for obtaining marine mesoscale eddy wave impedance based on deep learning is characterized by comprising the following steps:
step 1: establishing a sea water layer background sound velocity model and a sea water layer background density model, correspondingly adding a sound velocity model of a mesoscale vortex and a density model of a mesoscale vortex on the sea water layer background sound velocity model and the sea water layer background density model respectively to obtain a mesoscale vortex sound velocity model and a mesoscale vortex density model respectively,
the sound velocity model of the mesoscale vortex and the density model of the mesoscale vortex are obtained through presetting;
step 2: multiplying each of the plurality of different mesoscale vortex sound velocity models and each of the plurality of different mesoscale vortex density models respectively to obtain a plurality of mesoscale vortex wave impedance models,
the method comprises the following steps of obtaining a plurality of different mesoscale vortex sound velocity models and a plurality of different mesoscale vortex density models, wherein the specific implementation process comprises the following steps:
respectively rotating or translating the mesoscale vortex sound velocity model and the mesoscale vortex density model in the step 1, wherein the rotation angle is random, the translation distance is random, and a mesoscale vortex sound velocity model or a mesoscale vortex density model is correspondingly obtained in each rotation or translation, so that a plurality of different mesoscale vortex sound velocity models and a plurality of different mesoscale vortex density models are obtained;
and step 3: forward modeling is carried out on each group of mesoscale vortex models to obtain seismic reflection data, imaging processing is carried out on all the seismic reflection data respectively to obtain corresponding seismic reflection imaging sections,
wherein the set of mesoscale vortex models comprises the mesoscale vortex sound velocity model, the mesoscale vortex density model and the mesoscale vortex wave impedance model;
and 4, step 4: constructing a seismic reflection imaging section and a mesoscale vortex wave impedance model into tag data, wherein the tag data comprise target data and input data, the target data are the mesoscale vortex wave impedance model, and the input data are the seismic reflection imaging section;
and 5: building a deep learning model based on a neural network, dividing the label data in the step 4 into at least two parts, wherein one part is used for training the deep learning model, the other part is used for testing the deep learning model, and obtaining the finally trained deep learning model after the training and testing are qualified;
step 6: the actually collected seismic data is processed in the same way as the step 3 to obtain a seismic reflection imaging section of the wave impedance to be extracted,
and inputting the seismic reflection imaging section of the wave impedance to be extracted into the trained deep learning model for processing, wherein the result output by the deep learning model is the wave impedance of the mesoscale vortex, and the processing is finished.
2. The method for obtaining marine mesoscale eddy impedance based on deep learning of claim 1, wherein in step 3, in the forward modeling, the Rake wavelets with different frequencies are randomly selected as the seismic sources.
3. The method for obtaining the marine mesoscale eddy wave impedance based on the deep learning of claim 1, wherein in the forward simulation, a forward result is obtained by solving a sound wave equation of the mesoscale eddy model through numerical values, and the forward result is seismic reflection data, and the sound wave equation is as follows:
4. The method for obtaining the marine mesoscale eddy-wave impedance based on the deep learning of claim 1, wherein the neural network is obtained by combining a convolutional layer and a full connection layer, and the number of nodes of the input layer and the output layer in the neural network is Nz*Nx,NzNumber of sampling points in depth direction, NxIndicating the number of sample points in the horizontal lateral direction.
5. A processing terminal, characterized in that it comprises:
a memory for storing program instructions;
a processor for executing the program instructions to perform the steps of the method for obtaining mesoscale ocean eddy wave impedance based on deep learning according to any one of claims 1 to 4.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110078654.6A CN113312830B (en) | 2021-01-20 | 2021-01-20 | Method for obtaining marine mesoscale eddy wave impedance based on deep learning and processing terminal |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110078654.6A CN113312830B (en) | 2021-01-20 | 2021-01-20 | Method for obtaining marine mesoscale eddy wave impedance based on deep learning and processing terminal |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113312830A CN113312830A (en) | 2021-08-27 |
CN113312830B true CN113312830B (en) | 2022-02-25 |
Family
ID=77370589
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110078654.6A Active CN113312830B (en) | 2021-01-20 | 2021-01-20 | Method for obtaining marine mesoscale eddy wave impedance based on deep learning and processing terminal |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113312830B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114283128B (en) * | 2021-12-16 | 2024-04-26 | 中国人民解放军海军潜艇学院 | Ocean mesoscale vortex edge detection method and system based on multi-parameter threshold |
CN114417601B (en) * | 2022-01-18 | 2022-07-19 | 中国人民解放军国防科技大学 | Method for quickly estimating mesoscale eddy underwater sound velocity field based on satellite altimeter data |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102289594A (en) * | 2011-08-19 | 2011-12-21 | 中国科学院地理科学与资源研究所 | Algorithm for automatically identifying and reconstructing process of mesoscale ocean eddy |
CN110298280A (en) * | 2019-06-20 | 2019-10-01 | 上海海洋大学 | A kind of ocean eddy recognition methods based on MKL multiple features fusion |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110097075B (en) * | 2019-03-21 | 2023-04-18 | 国家海洋信息中心 | Deep learning-based marine mesoscale vortex classification identification method |
CN111767827A (en) * | 2020-06-28 | 2020-10-13 | 中国人民解放军国防科技大学 | Mesoscale vortex identification method based on deep learning |
CN112102325B (en) * | 2020-09-17 | 2021-11-09 | 中国科学院海洋研究所 | Ocean abnormal mesoscale vortex identification method based on deep learning and multi-source remote sensing data |
-
2021
- 2021-01-20 CN CN202110078654.6A patent/CN113312830B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102289594A (en) * | 2011-08-19 | 2011-12-21 | 中国科学院地理科学与资源研究所 | Algorithm for automatically identifying and reconstructing process of mesoscale ocean eddy |
CN110298280A (en) * | 2019-06-20 | 2019-10-01 | 上海海洋大学 | A kind of ocean eddy recognition methods based on MKL multiple features fusion |
Also Published As
Publication number | Publication date |
---|---|
CN113312830A (en) | 2021-08-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113312830B (en) | Method for obtaining marine mesoscale eddy wave impedance based on deep learning and processing terminal | |
CN112883564B (en) | Water body temperature prediction method and prediction system based on random forest | |
CN110031895B (en) | Multipoint geostatistical stochastic inversion method and device based on image stitching | |
CN112285768B (en) | High-frequency marine acoustic guided wave frequency dispersion analysis device and method | |
CN109521469B (en) | Regularization inversion method for elastic parameters of submarine sediments | |
CN113420440B (en) | Vortex identification method based on ocean vertical structure | |
CN111666529B (en) | Wave data processing and wave spectrum generating method | |
CN102109495A (en) | Method for classifying types of mixed seabed sediment based on multi-beam sonar technology | |
Maleika et al. | Interpolation methods and the accuracy of bathymetric seabed models based on multibeam echosounder data | |
CN107356666A (en) | A kind of extraction method and system of halmeic deposit parameters,acoustic | |
CN114049545B (en) | Typhoon intensity determining method, system, equipment and medium based on point cloud voxels | |
CN115859116A (en) | Marine environment field reconstruction method based on radial basis function regression interpolation method | |
CN115540828A (en) | Internal wave forecasting method based on wall sensor | |
Lin et al. | Merging multiple-partial-depth data time series using objective empirical orthogonal function fitting | |
CN116660996B (en) | Drifting type shallow sea local earth sound parameter prediction method based on deep learning | |
CN109188527B (en) | Method for rapidly establishing three-dimensional offshore bottom speed model in beach and shallow sea area | |
CN116861955A (en) | Method for inverting submarine topography by machine learning based on topography unit partition | |
CN113568041B (en) | Repeatability analysis method and system for time-lapse seismic three-dimensional towing cable acquired data | |
Ugwiri et al. | Edge sensor signal processing algorithms for earthquake early detection | |
CN115034115A (en) | Method for extracting third reflection interface ultrasonic echo weak signal in cased well by using deep learning | |
CN115408938A (en) | Method for inverting propagation speed of solitary wave in ocean by single-scene optical remote sensing image | |
TANG et al. | Application of LVQ neural network combined with the genetic algorithm in acoustic seafloor classification | |
CN112649848B (en) | Method and device for solving earthquake wave impedance by utilizing wave equation | |
CN113221651A (en) | Seafloor sediment classification method using acoustic propagation data and unsupervised machine learning | |
Watanabe et al. | Data assimilation of the stereo reconstructed wave fields to a nonlinear phase resolved wave model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |