Single set one-dimensional frequency energy density spectrum ratio measurement precision evaluation method and system

Abstract

The application discloses a kind of single group one-dimensional frequency energy density spectrum's ratio measurement precision evaluation method and system, belong to marine observation technical field, comprising: obtaining the data sample of sea wave one-dimensional frequency energy density spectrum of synchronous observation;The spectral feature value similarity of two synchronous observation spectra in data sample and spectral frequency energy density value sequence similarity are calculated;Based on spectral feature value similarity, the final comprehensive weight coefficient of spectral feature value is solved by least square method;Based on spectral frequency energy density value sequence similarity and the final comprehensive similarity of spectral feature value, the final comprehensive weight coefficient of one-dimensional frequency energy density spectrum is solved by least square method;The similarity evaluation value of two one-dimensional frequency energy density spectra to be evaluated is calculated using final comprehensive weight coefficient.The application can comprehensively and scientifically evaluate the similarity of two sea wave spectra by constructing double-layer evaluation model including spectral feature value similarity and spectral frequency energy density value sequence similarity, overcome the one-sidedness of traditional technology.

Description

Technical Field

[0001] This invention belongs to the field of marine observation technology, and in particular relates to a method and system for evaluating the comparative accuracy of a single set of one-dimensional frequency energy density spectra. Background Technology

[0002] In the field of ocean observation, the one-dimensional frequency energy density spectrum of ocean waves is a key element describing the distribution of energy within ocean waves at different frequencies, and it is of great significance for marine engineering, disaster prevention and mitigation, ship navigation, and wave energy development. To verify the performance of new wave spectrum observation instruments, it is usually necessary to conduct synchronous comparative tests with standard instruments.

[0003] However, there is currently a lack of a scientific, unified, and comprehensive evaluation method to quantify the consistency and similarity between two one-dimensional frequency energy density spectra observed by the instrument under test and the standard instrument. Existing technologies mostly use a single indicator (such as significant wave height or average period) for comparison, which cannot comprehensively reflect the differences in the entire spectrum, resulting in one-sided evaluation results and making it difficult to accurately assess the overall observation accuracy of the instrument under test. Summary of the Invention

[0004] This invention aims to address the lack of a comprehensive similarity evaluation method for two simultaneously observed one-dimensional frequency energy density spectra of ocean waves in the existing technology, and to provide a scientific, comprehensive and highly operable method for evaluating the accuracy of comparative measurements.

[0005] To achieve the above-mentioned objectives, the first objective of this invention is to provide a method for evaluating the comparative accuracy of a single set of one-dimensional frequency energy density spectra, comprising: Acquire samples of one-dimensional frequency energy density spectrum data of ocean waves from synchronous observation; Calculate the spectral feature value similarity and spectral frequency energy density value sequence similarity between two synchronously observed spectra in the data sample; Based on the spectral feature value similarity, the final comprehensive weight coefficient of the spectral feature value is solved by the least squares method; Based on the sequence similarity of the spectral frequency energy density values ​​and the final comprehensive similarity of the spectral feature values, the final comprehensive weighting coefficient of the one-dimensional frequency energy density spectrum is solved by the least squares method. The similarity evaluation value of the two one-dimensional frequency energy density spectra to be evaluated is calculated using the final comprehensive weighting coefficient.

[0006] Preferably, the spectral feature similarity includes nine parameters: the minimum to maximum ratio of effective wave height, the minimum to maximum ratio of average period, the minimum to maximum ratio of total wave energy, the minimum to maximum ratio of peak frequency, the minimum to maximum ratio of peak energy, the minimum to maximum ratio of spectral width, the minimum to maximum ratio of frequency spectrum centroid, the minimum to maximum ratio of total energy at 0.5 times the peak frequency, and the minimum to maximum ratio of frequency bandwidth at 0.5 times the peak energy.

[0007] Preferably, the similarity of the spectral frequency energy density value sequence includes the one-dimensional correlation coefficient and the one-dimensional cosine coefficient of the included angle between two synchronously observed one-dimensional frequency energy density spectra of ocean waves.

[0008] Preferably, the final comprehensive weighting coefficients for solving the spectral eigenvalues ​​include: A temporary comprehensive weighting coefficient for calculating the similarity of spectral feature values ​​is obtained by weighted summation of the similarity value itself, the difference between the similarity value and 1, and the difference between the similarity value and the minimum value. Calculate the temporary composite similarity of spectral eigenvalues; Establish a system of equations with spectral eigenvalue similarity as the variable and temporary comprehensive similarity as the constant; The final comprehensive weight coefficients are obtained by solving the system of equations using the least squares method.

[0009] Preferably, the final comprehensive weighting coefficients for solving the one-dimensional frequency energy density spectrum include: Calculate the final comprehensive similarity, one-dimensional correlation coefficient, and one-dimensional cosine coefficient of the included angle of the spectral eigenvalues ​​as temporary comprehensive weighting coefficients. Calculate the temporary comprehensive similarity of the one-dimensional frequency energy density spectrum; Establish a system of equations with the final comprehensive similarity of spectral eigenvalues, one-dimensional correlation coefficient, one-dimensional cosine coefficient of the included angle as variables, and temporary comprehensive similarity as a constant; The final comprehensive weight coefficients are obtained by solving the system of equations using the least squares method.

[0010] Preferably, the formula for calculating the similarity evaluation value is: ; Where RD is the similarity evaluation value, The first final overall weighting coefficient, The final comprehensive similarity of spectral eigenvalues. This is the second final comprehensive weighting coefficient. The correlation coefficient is a one-dimensional coefficient. This is the third final comprehensive weighting coefficient. It is the one-dimensional cosine coefficient of the included angle.

[0011] The second objective of this invention is to provide a system for evaluating the accuracy of a single set of one-dimensional frequency energy density spectra, comprising: The data acquisition module acquires one-dimensional frequency energy density spectrum data samples of ocean waves. The data samples include multiple sets of one-dimensional frequency energy density spectra observed simultaneously by the instrument under test and the standard instrument. The spectral feature similarity calculation module is used to calculate the similarity of spectral feature values ​​between two synchronously observed spectra in each data sample. The spectral feature value similarity includes the minimum and maximum ratio of effective wave height, the minimum and maximum ratio of average period, the minimum and maximum ratio of total wave energy, the minimum and maximum ratio of spectral peak frequency, the minimum and maximum ratio of spectral peak energy value, the minimum and maximum ratio of spectral width, the minimum and maximum ratio of frequency spectrum centroid, the minimum and maximum ratio of total energy at 0.5 times the spectral peak frequency, and the minimum and maximum ratio of frequency bandwidth at 0.5 times the spectral peak energy. The sequence similarity calculation module calculates the similarity of the spectral frequency energy density value sequences of two synchronously observed spectra in each data sample, including the one-dimensional correlation coefficient and the one-dimensional cosine coefficient of the included angle. The weight training module constructs a set of similarity equations based on the similarity between the spectral feature values ​​and the similarity between the spectral frequency energy density value sequences, and solves for the final comprehensive weight coefficients using the least squares method. The comprehensive evaluation value output module calculates the similarity evaluation value of the two one-dimensional frequency energy density spectra to be evaluated based on the final comprehensive weight coefficient.

[0012] Preferably, the weight training module includes a spectral feature weight solving unit and a comprehensive weight solving unit; The spectral feature weighting solution unit is configured as follows: A temporary comprehensive weighting coefficient for calculating the similarity of spectral feature values ​​is obtained by weighted summation of the similarity value itself, the difference between the similarity value and 1, and the difference between the similarity value and the minimum value. Calculate the temporary composite similarity of spectral eigenvalues; Establish a system of equations for spectral eigenvalue similarity with spectral eigenvalue similarity as the variable and temporary comprehensive similarity as the constant, and solve for the final comprehensive weight coefficient of the spectral eigenvalues; The comprehensive weight calculation unit is configured as follows: Calculate the final comprehensive similarity, one-dimensional correlation coefficient, and one-dimensional cosine coefficient of the included angle of the spectral eigenvalues ​​as temporary comprehensive weighting coefficients. Calculate the temporary comprehensive similarity of the one-dimensional frequency energy density spectrum; A set of one-dimensional frequency energy density spectrum similarity equations is established with the final comprehensive similarity of spectral eigenvalues, one-dimensional correlation coefficient, one-dimensional cosine coefficient of the included angle as variables, and temporary comprehensive similarity as a constant. The final comprehensive weight coefficient of the one-dimensional frequency energy density spectrum is then solved.

[0013] A third objective of this invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned method for evaluating the accuracy of a single set of one-dimensional frequency energy density spectra.

[0014] A fourth objective of this invention is to provide a computer program product, including a computer program that, when executed by a processor, implements the aforementioned method for evaluating the accuracy of a single set of one-dimensional frequency energy density spectra.

[0015] The advantages and positive effects of this application are: This invention constructs a two-layer evaluation model that includes the similarity of spectral feature values ​​and the similarity of spectral frequency energy density value sequences, which can comprehensively and scientifically evaluate the similarity of two wave spectra and overcome the one-sidedness of evaluation by a single index.

[0016] This invention adopts a data-driven approach, using a large number of samples to train and solve for the optimal weight coefficients, making the evaluation results more objective, stable and reliable.

[0017] The method disclosed in this invention is simple, easy to implement, and highly operable, making it suitable for large-scale application in comparative tests of ocean wave observation instruments. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart of the first embodiment of the present invention is shown; Figure 2 A model block diagram of the first embodiment of the present invention is shown; Figure 3 A flowchart of the unknown final comprehensive weighting coefficients in the first embodiment of the present invention is shown. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see Figure 1 The first embodiment, a method for evaluating the comparative accuracy of a single set of one-dimensional frequency energy density spectra, mainly includes: Acquire samples of one-dimensional frequency energy density spectrum data of ocean waves from synchronous observation; Calculate the spectral feature value similarity and spectral frequency energy density value sequence similarity between two synchronously observed spectra in the data sample; Based on the spectral feature value similarity, the final comprehensive weight coefficient of the spectral feature value is solved by the least squares method; Based on the sequence similarity of the spectral frequency energy density values ​​and the final comprehensive similarity of the spectral feature values, the final comprehensive weighting coefficient of the one-dimensional frequency energy density spectrum is solved by the least squares method. The similarity evaluation value of the two one-dimensional frequency energy density spectra to be evaluated is calculated using the final comprehensive weighting coefficient.

[0022] To better understand the technical solution of the present invention, a non-limiting description is provided below: Please see Figure 2 and Figure 3 , The similarity of a single set of one-dimensional frequency-energy density spectra consists of two parts: the similarity of spectral eigenvalues ​​and the similarity of the spectral frequency-energy density value sequences; where: The similarity of spectral eigenvalues ​​includes the ratio of the minimum to maximum effective wave height of two simultaneously observed one-dimensional frequency-energy density spectra of ocean waves. Average period minimum to maximum ratio Minimum to maximum ratio of total wave energy , the ratio of minimum to maximum spectral peak frequencies , the ratio of minimum to maximum peak energy values Spectral width minimum to maximum ratio Minimum-maximum ratio of the centroid of the frequency spectrum 0.5 times the ratio of the minimum to the maximum total energy of the spectral peak frequency 0.5 times the ratio of the minimum to the maximum frequency bandwidth of the spectral peak energy The final comprehensive similarity of spectral eigenvalues wait; The similarity of the spectral frequency energy density value sequences includes the one-dimensional correlation coefficient of two synchronously observed one-dimensional frequency energy density spectra of ocean waves. and the cosine coefficient of the one-dimensional angle Similarity evaluation value of two one-dimensional frequency energy density spectra. (Its value is between 0 and 1) and can be obtained using the following formula. The closer the value is to 1, the more similar the two one-dimensional frequency energy density spectra are. =0 indicates that they are completely dissimilar =1 indicates complete similarity; (1) (2) In the above formula, It is the fourth final comprehensive weighting coefficient. It is the fifth final comprehensive weighting coefficient. It is the sixth final comprehensive weighting coefficient. It is the seventh final comprehensive weighting coefficient. It is the eighth final comprehensive weighting coefficient. It is the ninth final comprehensive weighting coefficient. It is the tenth final comprehensive weighting coefficient. It is the eleventh final comprehensive weighting coefficient. It is the twelfth final comprehensive weighting coefficient. It is the first final comprehensive weighting coefficient. It is the second final comprehensive weighting coefficient. It is the third final comprehensive weighting coefficient.

[0023] Solve , , , , , , , , , , , The steps involving 12 unknown final composite weighting coefficients include: S1. Obtain a sample of one-dimensional frequency energy density spectrum data of ocean waves; Synchronous observations are obtained from data sources such as on-site sea trials, laboratory experiments, or computer simulations. ( ≥150 groups of one-dimensional frequency energy density spectra of ocean waves were used as data samples, and each group of data samples contained two synchronously observed one-dimensional frequency energy density spectra. S2. Preprocess one-dimensional frequency energy density spectrum data samples; The data samples of the one-dimensional frequency energy density spectrum are preprocessed to check for outliers. If outliers are found, the data samples are removed. S3. Solve for the similarity of spectral eigenvalues; Solve The similarity of spectral eigenvalues ​​among a set of one-dimensional frequency-energy density spectral data samples, i.e., the minimum-to-maximum ratio of spectral eigenvalues, is the first... ( =1,2,3… The minimum-to-maximum ratio of the spectral eigenvalues ​​of the one-dimensional frequency energy density spectrum of two synchronously observed data samples is respectively the minimum-to-maximum ratio of the effective wave height. The minimum and maximum ratio of the average period The minimum to maximum ratio of total wave energy The minimum to maximum ratio of spectral peak frequencies Minimum to maximum ratio of spectral peak energy values The minimum to maximum ratio of spectral width The minimum-to-maximum ratio of the centroids of the frequency spectrum The minimum to maximum ratio of the total energy at 0.5 times the peak frequency. The minimum to maximum ratio of 0.5 times the peak energy frequency bandwidth ; S4. Solve for the temporary comprehensive weighting coefficients of spectral eigenvalue similarity; Solve The temporary composite weighting coefficient for the similarity of spectral eigenvalues ​​of a group of one-dimensional frequency energy density spectral data samples, the first... ( =1,2,3… Temporary composite weighting coefficient for the similarity of spectral eigenvalues ​​of two synchronously observed one-dimensional frequency energy density spectra. , , , , , , , , It can be obtained from the following formula, where =1,2,3,4,5,6,7,8,9; The temporary comprehensive weighting coefficient consists of three parts. The first part of the temporary comprehensive weighting coefficient is shown below. In this part, the similarity value of the spectral feature value is used as a factor affecting the comprehensive similarity evaluation. The larger the similarity value of the spectral feature value, the greater its influence on the comprehensive similarity evaluation. Among them, the first temporary variable... Natural numbers from 1 to 9; ; The second part of the temporary comprehensive weighting coefficient is shown below. In this part, to prevent the comprehensive similarity evaluation results from becoming too concentrated and to reflect the magnitude gradient of the evaluation results, the difference between the similarity value of the spectral feature value and perfect similarity (similarity value of 1) is introduced as a factor affecting the comprehensive similarity evaluation. ; The third part of the temporary comprehensive weighting coefficient is shown below. In this part, in order to reflect the difference between the similarity of different spectral feature values, the difference from the minimum similarity value is introduced as a factor affecting the comprehensive similarity evaluation. , in for , , , , , , , , The minimum of 9 similarity values, among which, =1,2,3… ; =1,2,3,4,5,6,7,8,9; The temporary composite weighting coefficient for the similarity of spectral eigenvalues ​​is: , in, ,recommend =0.8, =0.1, =0.1.

[0024] S5. Solve for the temporary comprehensive similarity of spectral eigenvalues; Solve The temporary comprehensive similarity of the spectral eigenvalues ​​of a group of one-dimensional frequency-energy density spectral data samples, i.e., the temporary weighted sum coefficient of the minimum-maximum ratio, is the first... ( =1,2,3… The temporary weighted sum coefficient of the minimum-to-maximum ratio of the spectral eigenvalues ​​of two synchronously observed one-dimensional frequency energy density spectra is: The solution is obtained from the following formula: .

[0025] S6. Establish a system of similarity equations for spectral eigenvalues; Establish A set of similarity equations for the spectral eigenvalues ​​of a group of one-dimensional frequency-energy density spectral data samples, i.e., a set of equations for the minimum-maximum ratio, is derived from... The equations consist of the first, second, and third equations. ( =1,2,3… The equations are shown below, where, , , , , , , , , The final comprehensive weight coefficient to be solved. .

[0026] S7. Solve for the final comprehensive weight coefficient of spectral eigenvalue similarity; The least squares method is used to solve the system of similarity equations for spectral eigenvalues, and the final comprehensive weight coefficients of spectral eigenvalue similarity are obtained. , , , , , , , , .

[0027] S8. Solve for the final comprehensive similarity of the spectral eigenvalues; Solve The final comprehensive similarity of the spectral eigenvalues ​​of a group of one-dimensional frequency-energy density spectral data samples, i.e., the final weighted sum coefficient of the minimum-maximum ratio, is the first... ( =1,2,3… The final weighted sum coefficient of the minimum-to-maximum ratio of the spectral eigenvalues ​​of two synchronously observed one-dimensional frequency energy density spectra is: The solution is obtained from the following formula. .

[0028] S9. Solve for the similarity of the spectral frequency energy density value sequence; Solve The similarity of the spectral frequency energy density value sequence of a set of one-dimensional frequency energy density spectrum data samples, namely the one-dimensional correlation coefficient and the one-dimensional cosine coefficient of the one-dimensional frequency energy density spectrum, is given by the following formula: ( =1,2,3… The one-dimensional correlation coefficient of the one-dimensional frequency energy density spectra of the two synchronously observed groups is: The cosine coefficient of the one-dimensional included angle is ; S10. Solve for the temporary comprehensive weighting coefficients of the one-dimensional frequency energy density spectrum similarity; Solve The temporary comprehensive weighting coefficients for the one-dimensional frequency energy density spectrum similarity of a set of one-dimensional frequency energy density spectrum data samples include the temporary comprehensive weighting coefficients for the final comprehensive similarity of spectral feature values ​​and the temporary comprehensive weighting coefficients for the sequence similarity of spectral frequency energy density values. ( =1,2,3… The provisional comprehensive weighting coefficient for the final comprehensive similarity of the group spectrum feature values ​​is: The temporary comprehensive weighting coefficient of the one-dimensional correlation coefficient of the one-dimensional frequency energy density spectrum is: The temporary comprehensive weighting coefficient of the one-dimensional cosine coefficient of the one-dimensional frequency energy density spectrum is: It is obtained by the following formula, where the third temporary variable is... The values ​​are natural numbers from 1 to 3. The temporary comprehensive weighting coefficient consists of three parts. The first part of the temporary comprehensive weighting coefficient is shown below. In this part, the similarity value of the one-dimensional frequency energy density spectrum is used as a factor affecting the comprehensive similarity evaluation. The larger the similarity value, the greater its influence on the comprehensive similarity evaluation. Among them, the second temporary variable... Natural numbers from 1 to 3; ; The second part of the temporary comprehensive weighting coefficient is shown below. In this part, to prevent the comprehensive similarity evaluation results from becoming too concentrated and to reflect the magnitude gradient of the evaluation results, the difference between the similarity value of the one-dimensional frequency energy density spectrum and the perfect similarity (similarity value of 1) is introduced as a factor affecting the comprehensive similarity evaluation. ; The third part of the temporary comprehensive weighting coefficient is shown below. In this part, in order to reflect the difference between the similarity of different one-dimensional frequency energy density spectra, the difference from the minimum similarity value is introduced as a factor affecting the comprehensive similarity evaluation. , in, for , , The minimum of the three similarity values. =1,2,3… ; =1,2,3; The temporary comprehensive weighting coefficient for the similarity of the one-dimensional frequency energy density spectrum is: , in, ,recommend =0.8, =0.1, =0.1.

[0029] S11. Solve for the temporary comprehensive similarity of the one-dimensional frequency energy density spectrum; Solve Temporary composite similarity of one-dimensional frequency energy density spectrum data samples, the first ( =1,2,3… The temporary synthetic similarity of the one-dimensional frequency energy density spectra of the two synchronously observed groups is: It can be obtained from the following formula. .

[0030] S12. Establish a set of similarity equations for the one-dimensional frequency energy density spectrum; Establish A set of similarity equations for a group of one-dimensional frequency energy density spectrum data samples, which consists of... The similarity equations consist of the first, second, and third similarity equations. ( =1,2,3… The following are three similarity equations, where... , , The final comprehensive weight coefficient to be solved. . S13. Solve for the final comprehensive weighting coefficient of the one-dimensional frequency energy density spectrum similarity.

[0031] The least squares method is used to solve the similarity equations of the one-dimensional frequency energy density spectrum, and the final comprehensive weight coefficient of the similarity of the one-dimensional frequency energy density spectrum is obtained. , , .

[0032] The following example, using the one-dimensional frequency energy density spectrum simultaneously observed by two sets of wave buoys in a certain sea area, illustrates in detail the invention... , , , , , , , , , , , Specific implementation methods for the 12 unknown final comprehensive weighting coefficients.

[0033] S1. Obtain one-dimensional frequency energy density spectrum data samples of ocean waves, using 238 sets of one-dimensional frequency energy density spectra observed simultaneously by two sets of wave buoys in a certain sea area as data samples. Each set of one-dimensional frequency energy density spectrum data contains two one-dimensional frequency energy density spectra observed simultaneously by the two sets of wave buoys.

[0034] S2. Preprocess one-dimensional frequency energy density spectrum data samples. 238 sets of one-dimensional frequency energy density spectrum data samples were preprocessed and the data samples were checked. No outliers were found.

[0035] S3. Solve for the similarity of spectral eigenvalues. Calculate the similarity of spectral eigenvalues ​​for 238 sets of one-dimensional frequency-energy density spectrum data samples, i.e., the minimum-to-maximum ratio of spectral eigenvalues. As an implementation example, the minimum-to-maximum ratio of the first set of spectral eigenvalues ​​is given, including the minimum-to-maximum ratio of the significant wave height. =0.995173, the minimum and maximum ratio of the average period is =0.963255, the minimum to maximum ratio of total wave energy is =0.981853, the minimum to maximum ratio of spectral peak frequencies is =1, the minimum to maximum ratio of spectral peak energy values ​​is =0.970558, the minimum to maximum ratio of spectral width is =0.991638, the minimum-to-maximum ratio of the frequency spectrum centroid is =0.960364, the minimum to maximum ratio of the total energy at the 0.5 octave frequency is =0.714265, the minimum to maximum ratio of the frequency bandwidth of the 0.5 octave peak energy is =0.6. The solution process for other data samples is similar.

[0036] S4. Solve for the temporary composite weight coefficients of spectral eigenvalue similarity. This paper proposes a provisional comprehensive weighting coefficient for calculating the similarity of spectral eigenvalues ​​among 238 sets of one-dimensional frequency-energy density spectral data samples. As an implementation example, the provisional comprehensive weighting coefficient for the similarity of spectral eigenvalues ​​of two synchronously observed one-dimensional frequency-energy density spectra in the first set is provided. =0.8, =0.1, =0.1.

[0037] For sequence ( The minimum value among (1, 2, 3, 4, 5, 6, 7, 8, 9), where the first temporary variable... Natural numbers from 1 to 9; ; The similarity process is used to solve for the temporary comprehensive weight coefficients of other similarities, and the results are as follows: =0.111785; =0.112014; =0.112238; =0.111875; =0.112135; =0.111749; =0.108717; =0.107309; The solution process for other data samples is similar.

[0038] S5. Solve for the temporary comprehensive similarity of spectral eigenvalues. Solve for the temporary comprehensive similarity of the spectral eigenvalues ​​of 238 sets of one-dimensional frequency-energy density spectrum data samples, i.e., the temporary weighted sum coefficient of the minimum-maximum ratio. As an implementation example, the temporary comprehensive similarity of the first set of data samples is given: ; The solution process for other data samples is similar.

[0039] S6. Establish a system of similarity equations for spectral eigenvalues. This system comprises 238 equations for the spectral eigenvalues ​​of 238 sets of one-dimensional frequency-energy density spectrum data samples. The format of the first similarity equation is given as an implementation example: ; The solution process for other data samples is similar.

[0040] S7. Solve for the final comprehensive weighting coefficient of spectral eigenvalue similarity using the least squares method to solve the system of similarity equations for spectral eigenvalues, and obtain the final comprehensive weighting coefficient of spectral eigenvalue similarity. =0.227652; =0.086031; =0.046736; =0.096957; =0.091058; =0.112087; =0.155953; =0.082379; =0.092643.

[0041] S8. Solve for the final comprehensive similarity of spectral eigenvalues. Calculate the final comprehensive similarity of the spectral eigenvalues ​​for 238 sets of one-dimensional frequency-energy density spectral data samples, i.e., the final weighted sum coefficient of the minimum-maximum ratio. As an implementation example, the final comprehensive similarity of the spectral eigenvalues ​​for the first set of one-dimensional frequency-energy density spectral data samples is given: ; The solution process for other data samples is similar.

[0042] S9. Solve for the similarity of the spectral frequency energy density value sequences. Calculate the similarity of the spectral frequency energy density value sequences for 238 sets of one-dimensional frequency energy density spectral data samples, i.e., the one-dimensional correlation coefficient and the one-dimensional cosine coefficient of the one-dimensional frequency energy density spectrum. As an implementation example, the one-dimensional correlation coefficient of the first set of two synchronously observed one-dimensional frequency energy density spectra is given as follows: =0.980239, the one-dimensional cosine coefficient of the included angle is =0.985627. The solution process for other data samples is similar.

[0043] S10. Solve for the temporary comprehensive weighting coefficients of the one-dimensional frequency-energy density spectrum similarity. Solve for the temporary comprehensive weighting coefficients of the one-dimensional frequency-energy density spectrum similarity of 238 sets of one-dimensional frequency-energy density spectrum data samples, including the temporary comprehensive weighting coefficients of the final comprehensive similarity of spectral feature values ​​and the temporary comprehensive weighting coefficients of the sequence similarity of spectral frequency-energy density values. The temporary weighting coefficients for the first set of data samples are given as an implementation example. =0.8, =0.1, =0.1.

[0044] in for , , The minimum value of three similarity data, where the second temporary variable It is a natural number from 1 to 3.

[0045] ; A similar process was used to solve for the other temporary composite weight coefficients, and the results are as follows: =0.336827; =0.337787; The solution process for other data samples is similar. S11. Solve for the temporary comprehensive similarity of the one-dimensional frequency-energy density spectrum. Solve for the temporary comprehensive similarity of 238 sets of one-dimensional frequency-energy density spectrum data samples. As an implementation case, the temporary comprehensive similarity of the first set of two synchronously observed one-dimensional frequency-energy density spectra is given, which is obtained by the following formula: ; The solution process for other data samples is similar.

[0046] S12. Establish a set of similarity equations for one-dimensional frequency energy density spectra. A set of similarity equations for 238 sets of one-dimensional frequency energy density spectrum data samples is established. This set of equations consists of 238 similarity equations. The first similarity equation is given as an implementation example below: =0.961155; The solution process for other data samples is the same.

[0047] S13. Solve for the final comprehensive weighting coefficient of the one-dimensional frequency energy density spectrum similarity using the least squares method to solve the similarity equations of the one-dimensional frequency energy density spectrum and obtain the final comprehensive weighting coefficient: =0.319841; =0.279648; =0.399998.

[0048] Based on the solution already obtained above , , , , , , , , and , , The following model establishes a method for evaluating the comparative accuracy of a single set of one-dimensional frequency energy density spectra using 12 final comprehensive weighting coefficients. The similarity evaluation value for two synchronously observed one-dimensional frequency energy density spectra is also provided. (Its value is between 0 and 1) It is obtained from the following: ; .

[0049] A second embodiment provides a system for evaluating the accuracy of a single set of one-dimensional frequency energy density spectra, comprising: The data acquisition module acquires one-dimensional frequency energy density spectrum data samples of ocean waves. The data samples include multiple sets of one-dimensional frequency energy density spectra observed simultaneously by the instrument under test and the standard instrument. The spectral feature similarity calculation module is used to calculate the similarity of spectral feature values ​​between two synchronously observed spectra in each data sample. The spectral feature value similarity includes the minimum and maximum ratio of effective wave height, the minimum and maximum ratio of average period, the minimum and maximum ratio of total wave energy, the minimum and maximum ratio of peak frequency, the minimum and maximum ratio of peak energy, the minimum and maximum ratio of spectral width, the minimum and maximum ratio of frequency spectrum centroid, the minimum and maximum ratio of total energy at 0.5 times the peak frequency, and the minimum and maximum ratio of frequency bandwidth at 0.5 times the peak energy. The sequence similarity calculation module calculates the similarity of the spectral frequency energy density value sequences of two synchronously observed spectra in each data sample, including the one-dimensional correlation coefficient and the one-dimensional cosine coefficient of the included angle. The weight training module constructs a set of similarity equations based on the similarity between the spectral feature values ​​and the similarity between the spectral frequency energy density value sequences, and solves for the final comprehensive weight coefficients using the least squares method. The comprehensive evaluation value output module calculates the similarity evaluation value of the two one-dimensional frequency energy density spectra to be evaluated based on the final comprehensive weight coefficient.

[0050] The weight training module includes a spectral feature weight solving unit and a comprehensive weight solving unit; The spectral feature weighting solution unit is configured as follows: A temporary comprehensive weighting coefficient for calculating the similarity of spectral feature values ​​is obtained by weighted summation of the similarity value itself, the difference between the similarity value and 1, and the difference between the similarity value and the minimum value. Calculate the temporary composite similarity of spectral eigenvalues; Establish a system of equations for spectral eigenvalue similarity with spectral eigenvalue similarity as the variable and temporary comprehensive similarity as the constant, and solve for the final comprehensive weight coefficient of the spectral eigenvalues; The comprehensive weight calculation unit is configured as follows: Calculate the final comprehensive similarity, one-dimensional correlation coefficient, and one-dimensional cosine coefficient of the included angle of the spectral eigenvalues ​​as temporary comprehensive weighting coefficients. Calculate the temporary comprehensive similarity of the one-dimensional frequency energy density spectrum; A set of one-dimensional frequency energy density spectrum similarity equations is established with the final comprehensive similarity of spectral eigenvalues, one-dimensional correlation coefficient, one-dimensional cosine coefficient of the included angle as variables, and temporary comprehensive similarity as a constant. The final comprehensive weight coefficient of the one-dimensional frequency energy density spectrum is then solved.

[0051] In the third embodiment, a computer-readable storage medium stores a computer program that, when executed by a processor, implements the above-described method for evaluating the accuracy of a single set of one-dimensional frequency energy density spectra.

[0052] Fourth embodiment: A computer program product, including a computer program that, when executed by a processor, implements the above-described method for evaluating the accuracy of a single set of one-dimensional frequency energy density spectra.

[0053] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented, in whole or in part, as a computer program product, the computer program product includes one or more computer instructions. When the computer program instructions are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line, or wireless (e.g., infrared, wireless, microwave, etc.) means). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0054] The above description is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for evaluating the comparative accuracy of a single set of one-dimensional frequency energy density spectra, characterized in that, include: The one-dimensional frequency energy density spectrum data samples of ocean waves were obtained by the instrument under test and the standard instrument at the same time. The similarity of a single set of one-dimensional frequency energy density spectra consists of two parts: the similarity of spectral feature values ​​and the similarity of the spectral frequency energy density value sequence. Calculate the spectral eigenvalue similarity and spectral frequency energy density value sequence similarity between two synchronously observed spectra in the data sample. The spectral eigenvalue similarity includes nine parameters: minimum to maximum ratio of effective wave height, minimum to maximum ratio of average period, minimum to maximum ratio of total wave energy, minimum to maximum ratio of peak frequency, minimum to maximum ratio of peak energy, minimum to maximum ratio of spectral width, minimum to maximum ratio of frequency spectrum centroid, minimum to maximum ratio of 0.5 times peak frequency total energy, and minimum to maximum ratio of 0.5 times peak energy frequency bandwidth. The spectral frequency energy density value sequence similarity includes the one-dimensional correlation coefficient and the one-dimensional cosine coefficient of the included angle between the two synchronously observed one-dimensional frequency energy density spectra. Based on the spectral feature value similarity, the final comprehensive weight coefficient of the spectral feature value is solved by the least squares method; The final comprehensive similarity of spectral eigenvalues ​​is calculated based on the final comprehensive weight coefficients of the spectral eigenvalues. Based on the sequence similarity of the spectral frequency energy density values ​​and the final comprehensive similarity of the spectral feature values, the final comprehensive weighting coefficient of the one-dimensional frequency energy density spectrum is solved by the least squares method. The similarity evaluation value of the two one-dimensional frequency energy density spectra to be evaluated is calculated using the final comprehensive weighting coefficient.

2. The method for evaluating the comparative accuracy of a single set of one-dimensional frequency energy density spectra according to claim 1, characterized in that, The final comprehensive weighting coefficients for solving the spectral eigenvalues ​​include: Calculate the temporary comprehensive weighting coefficient for spectral eigenvalue similarity; Solve A temporary composite weighting coefficient for the similarity of spectral eigenvalues ​​of a group of one-dimensional frequency-energy density spectral data samples; the temporary composite weighting coefficient consists of three parts: The first part is: ; The second part is: ; The third part is: , in for , , , , , , , , The minimum of the 9 similarity values, among which, =1,2,3… ; =1,2,3,4,5,6,7,8,9; Natural numbers from 1 to 9; The temporary comprehensive weight coefficient is obtained by weighted summation of the three parts; Solve Temporary comprehensive similarity of spectral eigenvalues ​​of a group of one-dimensional frequency energy density spectral data samples; Establish a system of equations with spectral eigenvalue similarity as the variable and temporary comprehensive similarity as the constant; The final comprehensive weight coefficients are obtained by solving the system of equations using the least squares method.

3. The method for evaluating the comparative accuracy of a single set of one-dimensional frequency energy density spectra according to claim 1, characterized in that, The final comprehensive weighting coefficients for solving the one-dimensional frequency energy density spectrum include: Calculate the final comprehensive similarity, one-dimensional correlation coefficient, and one-dimensional cosine coefficient of the included angle of the spectral eigenvalues ​​as temporary comprehensive weighting coefficients. Calculate the temporary comprehensive similarity of the one-dimensional frequency energy density spectrum; Establish a system of equations with the final comprehensive similarity of spectral eigenvalues, one-dimensional correlation coefficient, one-dimensional cosine coefficient of the included angle as variables, and temporary comprehensive similarity as a constant; The final comprehensive weight coefficients are obtained by solving the system of equations using the least squares method.

4. The method for evaluating the comparative accuracy of a single set of one-dimensional frequency energy density spectra according to claim 1, characterized in that, The formula for calculating the similarity evaluation value is as follows: ； Where RD is the similarity evaluation value. The first final overall weighting coefficient, The final comprehensive similarity of spectral eigenvalues. This is the second final comprehensive weighting coefficient. The correlation coefficient is a one-dimensional coefficient. This is the third final comprehensive weighting coefficient. It is the one-dimensional cosine coefficient of the included angle.

5. A system for evaluating the accuracy of a single set of one-dimensional frequency energy density spectra, characterized in that, include: The data acquisition module acquires one-dimensional frequency energy density spectrum data samples of ocean waves simultaneously observed by the instrument under test and the standard instrument. The similarity of a single set of one-dimensional frequency energy density spectra consists of two parts: the similarity of spectral feature values ​​and the similarity of the spectral frequency energy density value sequence. The spectral feature similarity calculation module calculates the spectral feature value similarity between two synchronously observed spectra in the data sample. The spectral feature value similarity includes nine parameters: the minimum and maximum ratio of effective wave height, the minimum and maximum ratio of average period, the minimum and maximum ratio of total wave energy, the minimum and maximum ratio of peak frequency, the minimum and maximum ratio of peak energy, the minimum and maximum ratio of spectral width, the minimum and maximum ratio of frequency spectrum centroid, the minimum and maximum ratio of total energy at 0.5 times the peak frequency, and the minimum and maximum ratio of frequency bandwidth at 0.5 times the peak energy. The sequence similarity calculation module calculates the sequence similarity of spectral frequency energy density values ​​of two synchronously observed spectra in the data sample; the sequence similarity of spectral frequency energy density values ​​includes the one-dimensional correlation coefficient and the one-dimensional cosine coefficient of the included angle of the two synchronously observed wave one-dimensional frequency energy density spectra; The weight training module calculates the final comprehensive weight coefficients of the spectral feature values ​​based on the similarity of the spectral feature values ​​using the least squares method. The final comprehensive similarity of spectral eigenvalues ​​is calculated based on the final comprehensive weight coefficients of the spectral eigenvalues. Based on the sequence similarity of the spectral frequency energy density values ​​and the final comprehensive similarity of the spectral feature values, the final comprehensive weighting coefficient of the one-dimensional frequency energy density spectrum is solved by the least squares method. The comprehensive evaluation value output module uses the final comprehensive weighting coefficient to calculate the similarity evaluation value of the two one-dimensional frequency energy density spectra to be evaluated.

6. A computer program product, characterized in that, The method includes a computer program that, when executed by a processor, implements the method for evaluating the comparative accuracy of a single set of one-dimensional frequency energy density spectra as described in any one of claims 1-4.

7. A computer-readable storage medium storing a computer program, characterized in that, When executed by the processor, the program implements the method for evaluating the comparative accuracy of a single set of one-dimensional frequency energy density spectra as described in any one of claims 1-4.

Images