Multi-sea-condition-oriented wind wave and surge rapid separation and identification method and equipment
Through modal decomposition, feature recombination and spectrum recognition methods, combined with Gaussian and Lorentz models and genetic algorithms, the problem of lack of universality and rapidity of separation and recognition of wind and waves in the existing technology is solved, and the problem of separation and recognition of wind and waves is achieved. Adaptation and efficient identification of various sea conditions.
Patent Information
- Application Number
- CN202411468518.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-21
- Publication Date
- 2025-02-18
AI Technical Summary
The prior art lacks universality and rapidity in the separation and identification of wind and waves and swells, and it is difficult to adapt to a variety of sea conditions.
By obtaining the vertical displacement time series of ocean wave surfaces, modal decomposition is performed using Bayesian information criterion, the cosine distance of the modal components is calculated and hierarchical clustering is performed to obtain the feature recombination time series. Combining the current average wind power data, the spectrum is obtained using the fast Fourier transform, and the spectrum parts of wind and swells are identified through the Gaussian and Lorentz models, and the model parameters are optimized using genetic algorithms.
It realizes the rapid separation and identification of wind and waves and swells, adapts to a variety of sea conditions, and has strong practicality and applicability.
Smart Images
Figure CN119474929A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the fields of marine engineering and marine meteorology, and in particular to a method and device for rapid separation and identification of wind, wave and surge in multiple sea conditions. Background Art
[0002] During deep-sea resource exploration operations, we often face complex and changeable sea conditions, which puts forward new requirements for technologies in marine environmental perception, intelligent decision-making and control. In the process of marine resource exploration, due to limited monitoring methods and insufficient understanding of the characteristics of wind, wave and current, generation and propagation mechanisms, it is impossible to identify, predict and respond to environmental effects in the future in advance. The main effect of marine environmental factors such as wind, wave and current on exploration platforms / vessels is to produce six-degree-of-freedom offsets, among which the effect of waves is particularly obvious. The dynamic positioning system can ensure that the platform / vessel resists this influence, but its control input is related to the current wave state, often involving characteristics such as wave spectrum and wave parameters. Therefore, whether the wave state can be accurately identified is a prerequisite for ensuring stable control of the platform.
[0003] Waves are mainly divided into two categories: wind waves and swell waves. The action of wind and waves causes the platform / vessel to experience rapid acceleration changes and high-frequency motion, which is related to the fatigue and transient stability of the platform structure; swell waves have more obvious dynamic loads on the platform / vessel. Due to the obvious differences in the characteristics and impact mechanisms of wind waves and swell waves, it is necessary to identify the wave patterns and better design different controllers and control strategies for the system.
[0004] However, in the actual marine environment, especially in the deep sea, wind waves and swells usually coexist. The existing one-dimensional wind-swell separation method divides the spectrum of wind wave and swell components by observing and calculating the spectrum separation frequency. However, this method requires a full understanding of the wave spectrum and the characteristics of the wind and waves in the area. It is not only difficult to separate quickly, but also unable to clearly determine the changes in wind and swells.
[0005] At present, both wind wave and swell spectra are fitted by using empirical spectra such as JONSWAP (JP spectrum) or Pierson-Moskowitz (PM spectrum), but parameters often need to be adjusted according to local sea conditions. Although this spectrum-dependent model can better describe their amplitude-frequency characteristics, it cannot clearly describe the coexistence of wind and swell, and when double peaks appear, the spectrum-dependent model will be difficult to apply. Summary of the invention
[0006] The purpose of the present invention is to provide a method and device for rapid separation and identification of wind, wave and surge for multiple sea conditions in order to solve the problem that most of the current wind-surge separation and corresponding spectrum identification methods use modified empirical formulas and lack universality and rapidity.
[0007] The above-mentioned purpose of the present application is achieved through the following technical solutions:
[0008] S1: Obtain the time series of vertical displacement of ocean wave surface; perform modal decomposition on the time series of vertical displacement of ocean wave surface based on the Bayesian information criterion to obtain k modal components;
[0009] S2: Calculate the cosine distance of k modal components and perform hierarchical clustering to obtain the feature reorganization time series of the two groups;
[0010] S3: obtaining the current average wind data; obtaining the frequency spectrum of the characteristic reorganization time series through fast Fourier transform; analyzing the obtained two frequency spectra in combination with the current average wind data to obtain the wind wave component spectrum and the swell component spectrum;
[0011] S4: Identify the high-frequency parts of the wind wave component spectrum and the swell component spectrum through the Gaussian model; identify the low-frequency parts of the wind wave component spectrum and the swell component spectrum through the Lorentz model;
[0012] The actual values of the frequency spectrum of wind waves and swell waves are obtained; the parameters of the Gaussian model and the Lorentz model are optimized by combining the actual values of the frequency spectrum through a genetic algorithm to obtain the identified wind wave spectrum diagram and swell wave spectrum diagram.
[0013] This application adopts the above technical solution. Gaussian and Lorentz models are common functions used to describe the peak shape of data, and are widely used in physics and spectral statistics. The Lorentz model is generally a sharp peak, which is consistent with the low-frequency part of the spectrum. The Gaussian model is suitable for a relatively smooth distribution and can also better reflect their distribution in the high-frequency part. Based on the Gaussian and Lorentz models, the high- and low-frequency signals of the wave spectrum are respectively identified and superimposed, which can describe waves under various sea conditions.
[0014] Optionally, step S1 includes:
[0015] S11: Through variational mode decomposition, the vertical displacement time series of the ocean wave surface is decomposed to obtain k modal components:
[0016] The vertical displacement time series of the ocean wave surface is decomposed into k modal components, requiring the sum of the bandwidths of the center frequencies of each modal component to be minimal. The constrained variational model is as follows:
[0017]
[0018] where u k represents the kth mode; ω k represents the center frequency of the kth mode; δ(t) is the Dirac function, is the first derivative of time, and the * symbol indicates convolution;
[0019] u k (t) = A k (t)cos(φ k (t)), where A k (t) is the value of each mode u k The instantaneous amplitude varies with time; cos(φ k (t)) is a k (t) is the cosine function of the phase; is the Cauchy principal value; is a complex exponential function of frequency transformation; represents the square of the Euclidean norm;
[0020] st∑ k u k =f means constraining all u k The sum of is equal to the original ocean wave surface vertical displacement time series f;
[0021] S12: According to the Bayesian information criterion, determine the optimal number of decomposition modes k:
[0022] From 1 to 10, perform variational mode decomposition for each k and calculate the sum of squares of reconstruction errors E for each k. k , estimated likelihood function
[0023] Through the likelihood function Combined with the Bayesian information criterion, the BIC value under different k is as follows:
[0024]
[0025] Where n is the number of data points, k is the number of decomposed modes, is the maximum likelihood estimate, and
[0026] The k with the smallest BIC value is selected as the optimal number of decomposition modes.
[0027] Optionally, step S2 includes:
[0028] S21: normalize the k best decomposed modal components and calculate the cosine similarity to obtain the cosine distance;
[0029] The cosine similarity calculation formula of the first modal component and the second modal component is as follows:
[0030]
[0031] d=1-cos(θ)
[0032] Where n is the number of data points, X1 , X 2 They represent the first and second modal components obtained after decomposition, respectively; d represents the cosine distance;
[0033] S22: Hierarchical clustering is performed based on the cosine distance of each modal component to obtain two different sets of feature reorganization time series:
[0034] The cosine distances were converted into square matrices and clusters were merged by minimizing the total internal squared differences in the clustering process;
[0035] The maximum number of clusters was designed to be 2, and two types of characteristic mode groupings were obtained;
[0036] The two types of characteristic modes are grouped and superimposed to obtain two different sets of characteristic reorganization time series.
[0037] Optionally, step S3 includes:
[0038] Use fast Fourier transform to obtain the frequency domain signal of the feature reorganization time series;
[0039] The fast Fourier transform formula is as follows:
[0040]
[0041] Where N is the length of x(n), n is the time index, k is the frequency index, and j is the imaginary part;
[0042] Process the frequency domain signal and draw the spectrum of two sets of feature recombined time series;
[0043] Combined with the current average wind data, based on expert experience, the wind wave component spectrum and swell component spectrum corresponding to the spectra of the two sets of characteristic reorganized time series are identified;
[0044] The expert experience factors include: time series amplitude, wavelength, average wind force level, and frequency spectrum smoothness.
[0045] Optionally, step S4 includes:
[0046] The wind wave component spectrum and swell component spectrum are input into the Gaussian model and the Lorentz model, and the parameters of the Gaussian model and the Lorentz model are optimized based on the genetic algorithm to obtain the wind wave spectrum and swell spectrum:
[0047] Genetic algorithm is used for parameter optimization, and the optimization objectives are as follows:
[0048]
[0049] in, is the model fitting value, that is, the wind wave spectrum and surge wave spectrum after identification; f(x) is the actual value of the spectrum, and n is the total number of sample points.
[0050] Optionally, step S4 further includes:
[0051] The Gaussian model is as follows:
[0052]
[0053] Where μ is the mean of the Gaussian distribution and σ is the standard deviation;
[0054] This can be simplified into the following formula:
[0055]
[0056] Where a is the peak amplitude, b is the peak position, c is the peak width control parameter, and x represents the input wind wave component spectrum and swell component spectrum;
[0057] The Lorenz model is as follows:
[0058]
[0059] where x 0 The center position of the Lorentz distribution; the width at half-peak of γ;
[0060] Modify it into the following formula:
[0061]
[0062] where a 1 is the peak height, b 1 is the peak center position, c 1 is the full width at half maximum, and x represents the input wind wave component spectrum and swell component spectrum.
[0063] An electronic device comprises a processor, a memory, a user interface and a network interface, wherein the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes a method for rapid separation and identification of wind, wave and swell for multiple sea conditions.
[0064] A computer-readable storage medium stores instructions. When the instructions are executed, a method for rapid separation and identification of wind, wave and swell facing multiple sea conditions is executed.
[0065] The beneficial effects of the technical solution provided by this application are:
[0066] The wind and swell are separated by modal decomposition and feature recombination, and the frequency spectrum is identified based on Gaussian and Lorentz models respectively, and the parameters of the two models are optimized using genetic algorithms. This method can quickly separate and identify wind swells, and is adaptable to a variety of sea conditions, with strong practicality and applicability. BRIEF DESCRIPTION OF THE DRAWINGS
[0067] The present application will be further described below with reference to the accompanying drawings and embodiments, in which:
[0068] Figure 1 is a step diagram in an embodiment of the present application;
[0069] Figure 2 is a time series diagram of vertical displacement of the wave surface in the embodiment of the present application;
[0070] Figure 3 is a graph showing the change of BIC value with the modal number k in the embodiment of the present application;
[0071] Figure 4 It is a time-frequency domain result diagram of decomposition according to k=3 in the embodiment of the present application;
[0072] Figure 5 is a dendrogram after cluster visualization in an embodiment of the present application;
[0073] Figure 6 It is a characteristic recombination sequence diagram after two types of recombination in the embodiment of the present application;
[0074] Figure 7 are frequency spectra of the characteristic recombinant sequences in the embodiments of the present application after fast Fourier transformation;
[0075] Figure 8 is a surge spectrum fitting result diagram in an embodiment of the present application;
[0076] Fig. 9 is a wind and wave spectrum fitting result diagram in an embodiment of the present application;
[0077] Fig.10 It is a comparison diagram of the spectrum after the superposition of wind waves and swell waves in the embodiment of the present application and the actual wave spectrum;
[0078] Fig.11 is a comparison diagram of the spectrum after the superposition of wind waves and swell waves and the actual wave spectrum under another sea condition in an embodiment of the present application;
[0079] Fig.12 It is a schematic diagram of the structure of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION
[0080] In order to have a clearer understanding of the technical features, purposes and effects of the present application, the specific implementation methods of the present application are now described in detail with reference to the accompanying drawings.
[0081] The embodiments of the present application provide a method for rapid separation and identification of wind, wave and swell for multiple sea conditions.
[0082] Please refer to Figure 1 , Figure 1 This is a step diagram of a method for rapid separation and identification of wind, wave and swell in multiple sea conditions in an embodiment of the present application, including:
[0083] S1: Obtain the time series of vertical displacement of ocean wave surface; perform modal decomposition on the time series of vertical displacement of ocean wave surface based on the Bayesian information criterion to obtain k modal components;
[0084] S2: Calculate the cosine distance of k modal components and perform hierarchical clustering to obtain the feature reorganization time series of the two groups;
[0085] S3: obtaining the current average wind data; obtaining the frequency spectrum of the characteristic reorganization time series through fast Fourier transform; analyzing the obtained two frequency spectra in combination with the current average wind data to obtain the wind wave component spectrum and the swell component spectrum;
[0086] S4: Identify the high-frequency parts of the wind wave component spectrum and the swell component spectrum through the Gaussian model; identify the low-frequency parts of the wind wave component spectrum and the swell component spectrum through the Lorentz model;
[0087] The actual values of the frequency spectrum of wind waves and swell waves are obtained; the parameters of the Gaussian model and the Lorentz model are optimized by combining the actual values of the frequency spectrum through a genetic algorithm to obtain the identified wind wave spectrum diagram and swell wave spectrum diagram.
[0088] Step S1 includes:
[0089] S11: Through variational mode decomposition, the vertical displacement time series of the ocean wave surface is decomposed to obtain k modal components:
[0090] The vertical displacement time series of the ocean wave surface is decomposed into k modal components, requiring the sum of the bandwidths of the center frequencies of each modal component to be minimal. The constrained variational model is as follows:
[0091]
[0092] where u k represents the kth mode; ω k represents the center frequency of the kth mode; δ(t) is the Dirac function, is the first derivative of time, and the * symbol indicates convolution;
[0093] uk (t) = A k (t)cos(φ k (t)), where A k (t) is the value of each mode u k The instantaneous amplitude varies with time; cos(φ k (t)) is a k (t) is the cosine function of the phase; is the Cauchy principal value; is a complex exponential function of frequency transformation; represents the square of the Euclidean norm;
[0094] st∑ k u k =f means constraining all u k The sum of is equal to the original ocean wave surface vertical displacement time series f;
[0095] Specifically, each mode u k The demodulated signal After Hilbert transform and derivation, the obtained spectrum should be as concentrated as possible.
[0096] S12: According to the Bayesian information criterion, determine the optimal number of decomposition modes k:
[0097] From 1 to 10, perform variational mode decomposition for each k and calculate the sum of squares of reconstruction errors E for each k. k , estimated likelihood function
[0098] Through the likelihood function Combined with the Bayesian information criterion, the BIC value under different k is as follows:
[0099]
[0100] Where n is the number of data points, k is the number of decomposed modes, is the maximum likelihood estimate, and
[0101] The k with the smallest BIC value is selected as the optimal number of decomposition modes.
[0102] Specifically, by calculating the BIC values under different k respectively, a smaller BIC value indicates a better model fit, so the k under the minimum BIC is selected as the number of decomposition modes.
[0103] Specifically, select Figure 2 The variational mode decomposition of the vertical displacement time series of the wave surface under a certain sea condition is performed, and the value of the decomposition mode number k is determined based on the Bayesian information criterion. Figure 3It is the change of BIC value with the number of decomposition modes k after the sequence is decomposed. It can be seen that the optimal number of decomposition modes of variational mode decomposition is 3 at this time. Figure 4 is the result of decomposition based on k=3, Figure 4 (a) and Figure 4 (b) are their time-frequency domain diagrams respectively.
[0104] Step S2 includes:
[0105] S21: normalize the k best decomposed modal components and calculate the cosine similarity to obtain the cosine distance;
[0106] The cosine similarity calculation formula of the first modal component and the second modal component is as follows:
[0107]
[0108] d=1-cos(θ)
[0109] Where n is the number of data points, X 1 , X 2 They represent the first and second modal components obtained after decomposition, respectively; d represents the cosine distance;
[0110] S22: Hierarchical clustering is performed based on the cosine distance of each modal component to obtain two different sets of feature reorganization time series:
[0111] The cosine distances were converted into square matrices and clusters were merged by minimizing the total internal squared differences in the clustering process;
[0112] The maximum number of clusters was designed to be 2, and two types of characteristic mode groupings were obtained;
[0113] The two types of characteristic modes are grouped and superimposed to obtain two different sets of characteristic reorganization time series.
[0114] Specifically, according to the above formula Figure 4 By calculating the decomposed modal components, the cosine distance of each modal component can be obtained and converted into a square matrix. The clusters are merged by minimizing the total internal square difference in the clustering process, and the maximum number of clusters is designed to be 2. Finally, two types of characteristic modal groupings can be obtained. Figure 5 is a dendrogram after cluster visualization. By superimposing the modal components of these two categories, we can obtain two sets of different feature reorganization time series, such as Figure 6 shown.
[0115] Step S3 includes:
[0116] Use fast Fourier transform to obtain the frequency domain signal of the feature reorganization time series;
[0117] The fast Fourier transform formula is as follows:
[0118]
[0119] Where N is the length of x(n), n is the time index, k is the frequency index, and j is the imaginary part;
[0120] Process the frequency domain signal and draw the spectrum of two sets of feature recombined time series;
[0121] Combined with the current average wind data, based on expert experience, the wind wave component spectrum and swell component spectrum corresponding to the spectra of the two sets of characteristic reorganized time series are identified;
[0122] The expert experience factors include: time series amplitude, wavelength, average wind force level, and frequency spectrum smoothness.
[0123] Specifically, yes Figure 6 The characteristic reorganization time series is fast Fourier transformed to obtain Figure 7 The spectrum diagram shows that the average wind speed under this sea condition is 16.5m / s, which is level 7 wind force. Figure 6 neutron Figure 1 Large amplitude, long wavelength, and Figure 7 son Figure 1 It is smoother than 2, so the characteristic reorganization sequence 1 is the swell, and the other group is the wind wave component spectrum.
[0124] Step S4 includes:
[0125] The wind wave component spectrum and swell component spectrum are input into the Gaussian model and the Lorentz model, and the parameters of the Gaussian model and the Lorentz model are optimized based on the genetic algorithm to obtain the wind wave spectrum and swell spectrum:
[0126] Genetic algorithm is used for parameter optimization, and the optimization objectives are as follows:
[0127]
[0128] in, is the model fitting value, that is, the wind wave spectrum and surge wave spectrum after identification; f(x) is the actual value of the spectrum, and n is the total number of sample points.
[0129] Specifically, the wind wave spectrum diagram and the swell wave spectrum diagram are superimposed to obtain the wind wave swell wave spectrum diagram, which is compared and verified with the actual wind wave swell wave spectrum diagram to determine the optimal parameters of the Gaussian model and the Lorentz model under the current sea conditions.
[0130] Specifically, the Lorentz model is used to fit the swell part, and the Gaussian model is used to fit the wind wave. The genetic algorithm is used to optimize them respectively, and the following are obtained: Figure 8 , Fig. 9 The best fitting result is shown in the figure. It can be seen that the models can describe the original spectrum well. The current complete wave spectrum can be obtained by superimposing the wind wave and surge wave spectrum as follows: Fig.10 As shown in the figure, compared with the real spectrum, the two are basically consistent, which means that the method is feasible.
[0131] In order to better demonstrate that the proposed method can adapt to various sea conditions, a slightly complex sea condition with a double-peak spectrum is selected for verification. The average wind speed in this sea condition is 11.8m / s, which is level 6 wind force. Fig.11 It can be found that the wave spectrum presents a double-peak spectrum. The proposed methods can fit the spectrum well and realize wind surge separation.
[0132] Step S4 also includes:
[0133] The Gaussian model is as follows:
[0134]
[0135] Where μ is the mean of the Gaussian distribution and σ is the standard deviation;
[0136] This can be simplified into the following formula:
[0137]
[0138] Where a is the peak amplitude, b is the peak position, c is the peak width control parameter, and x represents the input wind wave component spectrum and swell component spectrum;
[0139] The Lorenz model is as follows:
[0140]
[0141] where x 0 The center position of the Lorentz distribution; the width at half-peak of γ;
[0142] Modify it into the following formula:
[0143]
[0144] where a 1 is the peak height, b 1 is the peak center position, c 1 is the full width at half maximum, and x represents the input wind wave component spectrum and swell component spectrum.
[0145] The present application also discloses an electronic device. Fig.12 , Fig.12The electronic device 500 may include: at least one processor 501 , at least one network interface 504 , a user interface 503 , a memory 505 , and at least one communication bus 502 .
[0146] The communication bus 502 is used to realize the connection and communication between these components.
[0147] The user interface 503 may include a display screen, and the optional user interface 503 may also include a standard wired interface or a wireless interface.
[0148] The network interface 504 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).
[0149] The present application also discloses a computer-readable storage medium storing a plurality of instructions suitable for loading by a processor to execute the above-mentioned method for rapid separation and identification of wind, wave and swell for multiple sea conditions.
[0150] The above are only exemplary embodiments of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification and the truth of practice, those skilled in the art will easily think of other embodiments of the present disclosure.
[0151] This application is intended to cover any variation, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the art not described in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.
Claims
1. A method for rapid separation and identification of wind, wave and swell in multiple sea conditions, characterized in that: The method comprises the following steps: S1: Obtain the time series of vertical displacement of ocean wave surface; perform modal decomposition on the time series of vertical displacement of ocean wave surface based on the Bayesian information criterion to obtain k modal components; S2: Calculate the cosine distance of k modal components and perform hierarchical clustering to obtain the feature reorganization time series of the two groups; S3: obtaining the current average wind data; obtaining the frequency spectrum of the characteristic reorganization time series through fast Fourier transform; analyzing the obtained two frequency spectra in combination with the current average wind data to obtain the wind wave component spectrum and the swell component spectrum; S4: Identify the high-frequency parts of the wind wave component spectrum and the swell component spectrum through the Gaussian model; identify the low-frequency parts of the wind wave component spectrum and the swell component spectrum through the Lorentz model; The actual values of the frequency spectrum of wind waves and swell waves are obtained; the parameters of the Gaussian model and the Lorentz model are optimized by combining the actual values of the frequency spectrum through a genetic algorithm to obtain the identified wind wave spectrum diagram and swell wave spectrum diagram.
2. A method for rapid separation and identification of wind, wave and swell in multiple sea conditions as claimed in claim 1, characterized in that: Step S1 includes: S11: Through variational mode decomposition, the vertical displacement time series of the ocean wave surface is decomposed to obtain k modal components: The vertical displacement time series of the ocean wave surface is decomposed into k modal components, requiring the sum of the bandwidths of the center frequencies of each modal component to be minimal. The constrained variational model is as follows: where u k represents the kth mode; ω k represents the center frequency of the kth mode; δ(t) is the Dirac function, is the first derivative of time, and the * symbol indicates convolution; u k (t) = A k (t)cos(φ k (t)), where A k (t) is the value of each mode u k The instantaneous amplitude varies with time; cos(φ k (t)) is a k (t) is the cosine function of the phase; is the Cauchy principal value; is a complex exponential function of frequency transformation; represents the square of the Euclidean norm; st∑ k u k =f means constraining all u k The sum of is equal to the original ocean wave surface vertical displacement time series f; S12: According to the Bayesian information criterion, determine the optimal number of decomposition modes k: From 1 to 10, perform variational mode decomposition for each k and calculate the sum of squares of reconstruction errors E for each k. k , estimated likelihood function Through the likelihood function Combined with the Bayesian information criterion, the BIC value under different k is as follows: Where n is the number of data points, k is the number of decomposed modes, is the maximum likelihood estimate, and The k with the smallest BIC value is selected as the optimal number of decomposition modes.
3. The method for rapid separation and identification of wind, wave and swell in multiple sea conditions as claimed in claim 1, characterized in that: Step S2 includes: S21: normalize the k best decomposed modal components and calculate the cosine similarity to obtain the cosine distance; The cosine similarity calculation formula of the first modal component and the second modal component is as follows: d=1-cos(θ) Where n is the number of data points, X1 and X2 represent the first and second modal components obtained after decomposition, respectively; d represents the cosine distance; S22: Hierarchical clustering is performed based on the cosine distance of each modal component to obtain two sets of different feature reorganization time series: The cosine distances were converted into square matrices and clusters were merged by minimizing the total internal squared differences in the clustering process; The maximum number of clusters was designed to be 2, and two types of characteristic mode groupings were obtained; The two types of characteristic modes are grouped and superimposed to obtain two different sets of characteristic reorganization time series.
4. The method for rapid separation and identification of wind, wave and swell in multiple sea conditions as claimed in claim 1, characterized in that: Step S3 includes: Use fast Fourier transform to obtain the frequency domain signal of the feature reorganization time series; The fast Fourier transform formula is as follows: Where N is the length of x(n), n is the time index, k is the frequency index, and j is the imaginary part; Process the frequency domain signal and draw the spectrum of two sets of feature recombined time series; Combined with the current average wind data, based on expert experience, the wind wave component spectrum and swell component spectrum corresponding to the spectra of the two sets of characteristic reorganized time series are identified; The expert experience factors include: time series amplitude, wavelength, average wind force level, and frequency spectrum smoothness.
5. The method for rapid separation and identification of wind, wave and swell in multiple sea conditions as claimed in claim 1, characterized in that: Step S4 includes: The wind wave component spectrum and swell component spectrum are input into the Gaussian model and the Lorentz model, and the parameters of the Gaussian model and the Lorentz model are optimized based on the genetic algorithm to obtain the wind wave spectrum and swell spectrum: Genetic algorithm is used for parameter optimization, and the optimization objectives are as follows: in, is the model fitting value, that is, the wind wave spectrum and surge wave spectrum after identification; f(x) is the actual value of the spectrum, and n is the total number of sample points.
6. A method for rapid separation and identification of wind, wave and swell in multiple sea conditions as claimed in claim 5, characterized in that: Step S4 also includes: The Gaussian model is as follows: Where μ is the mean of the Gaussian distribution and σ is the standard deviation; This can be simplified into the following formula: Where a is the peak amplitude, b is the peak position, c is the peak width control parameter, and x represents the input wind wave component spectrum and swell component spectrum; The Lorenz model is as follows: where x0 is the center position of the Lorentzian distribution; γ is the width at half peak; Modify it into the following formula: Where a1 is the peak height, b1 is the peak center position, c1 is the half-height full width, and x represents the input wind wave component spectrum and swell component spectrum.
7. An electronic device, characterized in that: The electronic device comprises a processor (501), a memory (505), a user interface (503) and a network interface (504), wherein the memory (505) is used to store instructions, the user interface (503) and the network interface (504) are used to communicate with other devices, and the processor (501) is used to execute the instructions stored in the memory (505) so that the electronic device executes the method according to any one of claims 1 to 6.
8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed by a computer, the method according to any one of claims 1 to 6 is executed.