Wave spectrum model selection method

Through frequency domain analysis of sea area wave sequences and comprehensive evaluation index calculation, wave spectrum models that meet the actual sea area are screened, solving the applicability and error problems of wave spectrum models in the existing technology, and achieving more accurate wave spectrum description and model choose.

CN119004254APending Publication Date: 2024-11-22CHINA HYDROELECTRIC ENGINEERING CONSULTING GROUP CHENGDU RESEARCH HYDROELECTRIC INVESTIGATION DESIGN AND INSTITUTE
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411039936.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-31
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The existing wave spectral models are mostly summarized in field observation data in specific sea areas and cannot be widely used in all marine environments. The selection method based on correlation coefficients cannot effectively represent the local error of spectral characteristics.

Method used

Frequency domain analysis is performed through wave sequences in the measured sea area, the measured spectrum is obtained, and the wave spectrum model is screened based on the spectrum width parameters and spectral peak period, comprehensive evaluation index is calculated, and the model closest to 1.0 is selected as the optimal model, combined with spectral model The proportion of feature parameters and fitting results is determined.

Benefits of technology

Accurate modeling and positioning of wave spectrum is achieved, a wave spectrum model that conforms to the actual sea area is provided, and the applicability and accuracy of the model is improved, and new models can be added for verification in the subsequent period.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119004254A_ABST
    Figure CN119004254A_ABST
Patent Text Reader

Abstract

The invention relates to the field of ocean engineering, and provides a wave spectrum model selection method in order to improve the accuracy of wave spectrum model selection, the wave spectrum models are screened by comparing spectrum width parameters and spectrum peak periods between a fitting spectrum and an actually measured spectrum, and on the basis that the spectrum width parameters of the fitting spectrum are close to the spectrum width parameters of the actually measured spectrum, a wave spectrum model is selected. If the spectrum peak period of the fitting spectrum is equal to the spectrum peak period of the actually measured spectrum, the wave spectrum model is an optimal model; otherwise, calculating a comprehensive evaluation index based on the spectrum width parameter and the spectrum peak period, and taking the wave spectrum model with the comprehensive evaluation index closest to 1.0 as the optimal model. The spectrum width parameter and the spectrum peak period represent whether the wave spectrum model can meet the actually measured wave spectrum in the wave spectrum shape and the spectrum peak position respectively, the wave spectrum shaping and positioning effects are achieved, and the selected model is more accurate based on the wave spectrum shaping and positioning effects.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of marine engineering, and in particular to a wave spectrum model selection method. Background Art

[0002] Waves are the most important dynamic factor in the ocean and the main external load borne by ships and marine engineering structures. Waves in the ocean are a complex random process. It is impossible to fully and systematically describe waves only by statistical parameters and wave characteristics from the time domain perspective. Since the early 1950s, using wave spectra to describe waves from the frequency domain perspective has become an important research approach. At present, researchers have proposed many wave spectrum models based on different environmental conditions and wave types, hoping to qualitatively describe complex wave processes through simple mathematical models. However, most of the existing wave spectrum models are obtained by summarizing data and laws based on field observation data in a specific sea area, and cannot be widely used in all marine environments. Therefore, it is very important to select a wave spectrum suitable for the application sea area from the many existing wave spectrum models. At present, the selection method based on correlation coefficient is mostly adopted. The correlation coefficient evaluation system can only represent the overall error of the curve, but the overall error will cover up the local error of the most concerned spectral characteristics. Summary of the invention

[0003] In order to improve the accuracy of wave spectrum model selection, the present application provides a wave spectrum model selection method.

[0004] The technical solution adopted by the present invention to solve the above problems is:

[0005] The wave spectrum model selection method includes:

[0006] Step 1: Measure the wave sequence of the sea area to be analyzed;

[0007] Step 2, perform frequency domain analysis on the wave sequence to obtain the measured spectrum;

[0008] Step 3, obtaining the fitting spectrum of each wave spectrum model in the model library based on the measured spectrum;

[0009] Step 4, calculating the spectrum width parameters and spectrum peak periods corresponding to the measured spectrum and each fitted spectrum;

[0010] Step 5, select the wave spectrum model whose spectral width parameter in the fitted spectrum is closest to the spectral width parameter of the measured spectrum, and determine whether its spectral peak period is equal to the spectral peak period of the measured spectrum. If so, take the wave spectrum model as the optimal model, otherwise proceed to step 6;

[0011] Step 6: Calculate the comprehensive evaluation index vT p , the wave spectrum model with the most comprehensive evaluation index close to 1.0 is taken as the optimal model; νT p =K1×(νfit / v best )+K2×(T pfit / T pbest ), where K1 and K2 represent the weights of the spectrum width parameter and the spectrum peak period, respectively, K1+K2=1, ν fit is the spectral width parameter corresponding to the fitted spectrum, T pfit is the peak period of the fitted spectrum, T pbest is the peak period corresponding to the measured spectrum, ν best Spectral width parameter corresponding to the measured spectrum.

[0012] Furthermore, the step 1 obtains multiple wave sequences; multiple wave spectrum models are used to fit the wave sequences to obtain multiple fitting spectrum results, and the number of times the wave spectrum model corresponding to each fitting spectrum is selected as the optimal model is counted, and the model with the largest proportion among all the optimal models is used as the final model.

[0013] Furthermore, the method further includes step 7, calculating the wave spectrum area based on the fitting spectrum corresponding to the optimal model to obtain the error between the fitting wave height and the measured wave height.

[0014] Furthermore, the step 2 uses fast Fourier transform to perform frequency domain analysis.

[0015] Furthermore, the step 3 specifically includes: performing curve fitting based on the measured spectrum using the nonlinear least square method to obtain the fitting spectrum of each wave spectrum model in the model library.

[0016] Furthermore, the model library includes: the Brookfield spectrum model, the ISSC spectrum model, the ITTC spectrum model, the BM spectrum model, the JONSWAP spectrum model and the Wallops spectrum model.

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] This application selects the wave spectrum model by comparing the spectral width parameters and spectral peak period between the fitted spectrum and the measured spectrum. On the basis that the spectral width parameters of the fitted spectrum are close to the spectral width parameters of the measured spectrum, if the spectral peak period of the fitted spectrum is equal to the spectral peak period of the measured spectrum, then the wave spectrum model is the optimal model; otherwise, the comprehensive evaluation index is calculated based on the spectral width parameters and the spectral peak period, and the wave spectrum model with the comprehensive evaluation index closest to 1.0 is taken as the optimal model. The spectral width parameters and spectral peak period respectively represent whether the wave spectrum model can meet the measured wave spectrum in terms of spectral shape and spectral peak position, thereby achieving the shaping and positioning of the wave spectrum. The model selected based on this is more accurate.

[0019] This application has strong flexibility and scalability, and a new wave spectrum model can be added as a fitting basis function in the future as a means of testing whether the newly developed wave spectrum model is suitable for actual sea areas.

[0020] This application can provide an input wave spectrum that conforms to the actual sea area and is relatively accurate for ship design and ocean wave environment assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 Flowchart of the method for wave spectrum model selection;

[0022] Figure 2 It is the measured wave sequence and the corresponding wave spectrum;

[0023] Figure 3 This is the optimal wave spectrum model selection and proportion diagram. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0025] like Figure 1 As shown in FIG. 1 , the wave spectrum model selection method includes:

[0026] Step 1: Measure the wave sequence in the sea area to be analyzed.

[0027] The example data comes from wave data obtained from actual observations in a certain sea area. The details of the data are shown in Table 1.

[0028] Table 1 Measured wave data

[0029] Data Types Sampling frequency Sampling time Total amount of data Wave process line 4Hz 4096s 1355

[0030] Step 2: Perform frequency domain analysis on the wave sequence to obtain the measured spectrum.

[0031] The measured wave process line η(t) is processed by fast Fourier transform, see formula (1), and the measured wave time process line is converted into a frequency domain result that changes with frequency. The time domain wave sequence and its corresponding frequency domain result are shown in the attached figure. Figure 2 ;

[0032]

[0033] in,

[0034]

[0035] Let N represent the total number of samples, Δt represent the sampling time interval, then the total sampling duration is NΔt. Consider NΔt as a complete cycle (or a multiple of the cycle), and the summation symbol represents the sum of the discrete points in the entire cycle, that is, 2l = NΔt. In this way, a measured time process sequence is approximated to the sum of a finite number of cosine functions and a constant.

[0036] Step 3: Obtain the fitting spectrum of each wave spectrum model in the model library based on the measured spectrum.

[0037] The wave spectrum expressions and coefficients in the model library are shown in formula (2) and Table 2. The actual wave spectrum is fitted using the nonlinear least squares method, see formula (3), and the fitting spectrum of each model is obtained.

[0038] S(f)=A f- 5 exp(-B f- 4 )(2)

[0039] Table 2 Expressions of coefficients A and B in different wave spectrum models

[0040]

[0041] in, and Represent the average wave height and average period respectively; H s and T s represents the effective wave height and effective period. For ITTC spectrum, α=0.0081, For the JONSWAP spectrum model, the expression includes the peak enhancement factor γ, where the coefficient β J The expressions for the peak rise terms are:

[0042]

[0043] Among them, f p is the spectrum peak frequency. When f≤fp, σ=0.07; when f>fp, σ=0.09.

[0044] For the Wallops spectrum model, the expression includes the high-frequency tail spectrum parameter m, the coefficient β w The expression is:

[0045]

[0046] Among them, x is a vector consisting of the parameters in the model, and e(x) is the residual vector.

[0047] Step 4: Calculate the spectral width parameters and spectral peak periods corresponding to the measured spectrum and each fitted spectrum.

[0048] The quality of the fitting results mainly focuses on the degree of agreement between the fitted spectrum and the measured spectrum. For the wave spectrum, we mainly focus on the parameters that can represent the characteristics of the spectrum, including the spectrum width parameter ν and the spectrum peak period Tp, and the calculation method is shown in formulas (4)-(6). They respectively represent whether the wave spectrum model can meet the measured wave spectrum in terms of spectral shape and spectral peak position, and realize the shaping and positioning of the wave spectrum. The model selected based on this will be more accurate.

[0049]

[0050] Among them, m r Represents the r-order moment of the wave spectrum. The moment is defined as follows:

[0051]

[0052] T p =1 / f p (6)

[0053] Step 5, select the wave spectrum model whose spectral width parameter in the fitted spectrum is closest to the spectral width parameter of the measured spectrum, and determine whether its spectral peak period is equal to the spectral peak period of the measured spectrum. If so, take the wave spectrum model as the optimal model, otherwise proceed to step 6;

[0054] Step 6: Calculate the comprehensive evaluation index νT p , and the wave spectrum model with the most comprehensive evaluation index close to 1.0 is taken as the optimal model; νT p =K1×(ν fit / v best )+K2×(T pfit / T pbest ), where K1 and K2 represent the weights of the spectrum width parameter and the spectrum peak period respectively, K1+K2=1, and the specific value is set according to actual needs and is not limited here; ν fit is the spectral width parameter corresponding to the fitted spectrum, T pfit is the peak period of the fitted spectrum, T pbest is the peak period corresponding to the measured spectrum, ν best Spectral width parameter corresponding to the measured spectrum.

[0055] Furthermore, the method also includes step 7, calculating the wave spectrum area based on the fitting spectrum corresponding to the optimal model to obtain the error between the fitting wave height and the measured wave height, so as to provide an error reference for the user.

[0056] After the optimal wave spectrum model is obtained by selecting the spectral characteristic parameters, it is verified by the spectral area of ​​the fitted spectrum and the measured spectrum. The spectral area S represents the sum of the wave energy. The calculation method is shown in formula (7), which represents the performance of the fitting result in terms of total wave energy. The relationship between the wave spectrum area S and the wave effective wave height Hs is shown in formula (8).

[0057]

[0058] The model error selected based on a single measured spectrum is large, so we can set several groups of measured spectra, perform statistical analysis on all the results to obtain the number of times each wave spectrum model is selected as the optimal solution, and then select the wave spectrum expression suitable for the specific sea area. Figure 3 As shown in the figure, the Wallops spectrum is selected as the optimal solution the most times, so the Wallops spectrum model is determined as the final model.

Claims

1. A wave spectrum model selection method, characterized in that: include: Step 1: Measure the wave sequence of the sea area to be analyzed; Step 2, perform frequency domain analysis on the wave sequence to obtain the measured spectrum; Step 3, obtaining the fitting spectrum of each wave spectrum model in the model library based on the measured spectrum; Step 4, calculating the spectrum width parameters and spectrum peak periods corresponding to the measured spectrum and each fitted spectrum; Step 5, select the wave spectrum model whose spectrum width parameter in the fitting spectrum is closest to the spectrum width parameter of the measured spectrum, and determine whether its spectrum peak period is equal to the spectrum peak period of the measured spectrum. If so, take the wave spectrum model as the optimal model, otherwise proceed to step 6; Step 6: Calculate the comprehensive evaluation index vT p , the wave spectrum model with the most comprehensive evaluation index close to 1.0 is taken as the optimal model; νT p =K1×(ν fit / v best )+K2×(T pfit / T pbest ), where K1 and K2 represent the weights of the spectrum width parameter and the spectrum peak period, respectively, K1+K2=1, ν fit is the spectral width parameter corresponding to the fitted spectrum, T pfit is the peak period of the fitted spectrum, T pbest is the peak period corresponding to the measured spectrum, ν best Spectral width parameter corresponding to the measured spectrum.

2. The wave spectrum model selection method according to claim 1, characterized in that: The step 1 obtains multiple wave sequences; multiple wave spectrum models are used to fit the wave sequences to obtain multiple fitting spectrum results, and the number of times the wave spectrum model corresponding to each fitting spectrum is selected as the optimal model is counted, and the model with the largest proportion among all the optimal models is used as the final model.

3. The wave spectrum model selection method according to claim 1, characterized in that: The method further includes step 7, calculating the wave spectrum area based on the fitting spectrum corresponding to the optimal model to obtain the error between the fitting wave height and the measured wave height.

4. The wave spectrum model selection method according to claim 1, characterized in that: The step 2 uses fast Fourier transform to perform frequency domain analysis.

5. The wave spectrum model selection method according to claim 1, characterized in that: The step 3 specifically includes: performing curve fitting based on the measured spectrum using the nonlinear least square method to obtain the fitting spectrum of each wave spectrum model in the model library.

6. The wave spectrum model selection method according to any one of claims 1 to 5, characterized in that: The model library includes: the Brookfield spectrum model, the ISSC spectrum model, the ITTC spectrum model, the BM spectrum model, the JONSWAP spectrum model and the Wallops spectrum model.

Citation Information

Cited By

  • Fitting method of wave spectrum mathematical expression of constrained secondary gravity wave generated by short wave group

    CN120354632A

  • Method, medium and system for calculating ocean wave spectrum parameters of ocean digital twinning

    CN120597727A

  • A method, medium, and system for ocean digital twinning sea wave spectrum parameter calculation

    CN120597727B