Star line-of-sight velocity calculation method, terminal device and storage medium

Abstract

The application discloses a star radial velocity calculation method, a terminal device and a storage medium. After obtaining multiple pieces of spectrum data of a target star, each piece of spectrum is divided into multiple sections, and normalization processing and sampling are performed on the sections. A Gaussian process regression model of the spectrum is established by using the sampled spectrum data. Absorption lines in the spectrum are extracted, the absorption lines are divided into multiple groups according to the relative depth of the absorption lines, the weight of each group relative to the original spectrum is calculated. Based on the spectrum model, the weight value of the absorption lines is combined to calculate the radial velocity difference between different spectrums. The radial velocity of the star is obtained by using the radial velocity difference between different spectrums. The application improves the calculation accuracy of the radial velocity of the star and reduces the calculation amount in the spectrum modeling process.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of exoplanet detection, and particularly to a method for calculating radial velocity of a star, a terminal device and a storage medium. BACKGROUND

[0002] The radial velocity method is a new exoplanet detection technique. Researchers calculate the radial velocity of a target star by observing its spectrum, and then determine whether an exoplanet exists according to the radial velocity of the star.

[0003] A typical method for calculating radial velocity is the cross-correlation function (CCF) algorithm. The template spectrum is moved multiple times, and the cross-correlation coefficient of the moved template spectrum and the observed spectrum is calculated. The radial velocity of the target star is then obtained according to the change in the cross-correlation coefficient. The CCF algorithm can achieve relatively accurate results, but this scheme still has some defects. First, there are differences between different stars, and their spectra are not completely consistent with the template spectrum, which means that there is a certain error in calculating the radial velocity using the CCF algorithm. Second, the star surface activity will change the star spectrum, and the change in the spectrum will also reduce the accuracy of the radial velocity calculation. Third, the calculation result of the CCF algorithm is mainly affected by the signal-to-noise ratio of the observed spectrum and the quality of the template spectrum. Multiple observations of the target star cannot improve the accuracy of each calculation result because the observation time is different each time. To address the above problems, researchers have proposed other solutions. Yizhen Ping selected a spectral segment that is less affected by noise and calculated the radial velocity of M-type stars using the CCF algorithm. Cao Hui calculated the radial velocity of the target star according to the shift of several star spectral lines. Foreign researchers designed high-precision spectrometers such as SOPHIE, HARPS, and ESPRESSO to improve the quality of the observed spectrum and obtain more accurate radial velocity values.

[0004] There are still some problems in the existing radial velocity algorithm. First, the flux values at different wavelengths are not completely independent for one spectrum. In other words, after the flux value of a point and its uncertainty are known, the uncertainty of the flux values around it can be reduced. For a high-resolution spectrometer, there are a large number of sample points for each spectrum, and these data have certain redundancy, making the processing of the spectral data very cumbersome. Second, the generation of a template spectrum by an algorithm or the use of a high signal-to-noise ratio spectrum as a template spectrum cannot completely solve the problem of the inconsistency between the template spectrum and the observed spectrum, which will increase the error of the final result. Third, directly comparing the spectra of a star at two times cannot accurately obtain the radial velocity change value of the star. Since the spectrum of a star is affected by factors such as stellar activity and noise, directly comparing two spectra to obtain the radial velocity difference will have a large error, and the radial velocity difference with a large error will affect the analysis of the radial velocity change of the star. Fourth, considering that absorption lines with different relative depths contain different radial velocity information, in order to fully extract the information, the absorption lines in the spectrum need to be weighted, but without relying on the template spectrum, the star spectrum cannot be directly weighted, and a reasonable weight distribution algorithm needs to be designed. SUMMARY

[0005] The technical problem to be solved by the present application is to provide a star radial velocity calculation method, a terminal device and a storage medium to improve the accuracy of star radial velocity calculation and simplify the calculation process.

[0006] To solve the above technical problems, the technical scheme adopted by the present application is as follows: a star radial velocity calculation method, comprising the following steps:

[0007] S1, obtaining S spectra, dividing each spectrum into T segments, and preprocessing the spectrum data of each segment;

[0008] S2, constructing a Gaussian process regression model for each segment of the spectrum using the preprocessed spectrum data of each segment, calculating the mean of the flux values at different wavelengths of each segment of the spectrum and the covariance between each flux value;

[0009] S3, extracting the absorption lines in each segment of the spectrum according to the mean of the flux values at different wavelengths of each segment of the spectrum, grouping the absorption lines with the same relative depth into a group to form a partial spectrum, and calculating the weight of each partial spectrum relative to the original spectrum;

[0010] S4, calculating the radial velocity difference and variance between the partial spectra obtained by the same threshold segmentation of the same segment of different spectra;

[0011] S5, repeating step S4, calculating multiple radial velocity differences and variances by using multiple partial spectra of the same section of the two spectra, and performing weighted summation on the multiple radial velocity differences by using the weights in step S3 to calculate the radial velocity difference and variance between the same section of the two spectra;

[0012] S6, repeating steps S4-S5, calculating multiple radial velocity differences and variances according to multiple section spectral data of the two spectra, and calculating the radial velocity difference between the two spectra;

[0013] S7, obtaining the radial velocity of the spectrum according to the radial velocity difference between the spectrum and other spectra.

[0014] The present application uses a Gaussian process regression algorithm to process spectral data, reduces the error of the spectral data, and obtains more accurate flow values. After the spectral modeling is realized, the radial velocity difference of the star at different times is calculated, and more accurate results are obtained. Based on the periodicity of the signal, the radial velocity difference of the star at different times is used to solve the radial velocity value, so that more accurate radial velocity of the star at different times can be obtained.

[0015] The specific implementation process of step S1 includes:

[0016] Calculation Wherein, λ t ' represents the wavelength of the t section of the i spectrum, f i (λ t ') represents the flow value of the t section of the i spectrum, f i (λ t ') represents the minimum value of f i (λ′ t ) min , the maximum value is f i (λ′ t ) max , σ′ i (λ′ t ) represents the error estimate value of each flow value, f i (λ′ t ) represents the normalized flow value of the t section of the i spectrum, σ i (λ′ t ) represents the error estimate value of f i (λ′ t );

[0017] Sampling the normalized t section of the i spectrum, the sampled spectral wavelength is λ t , the flow value is f i (λ t ), and the error estimate value is σ i (λ t ).

[0018] In step S2, for the t-th segment of the i-th spectrum, the Gaussian process regression model is constructed as follows:

[0019] The observed spectrum is represented as: i (λ t ) = y i (λ t ) + n i (λ t ), n i (λ t ) represents Gaussian noise, y i (λ t ) is unknown in form, and a Gaussian process prior is given to it, i.e., y i (λ t ) ~ GP(m i (λ t ), K i (λ t , λ t )), the prior distribution of f i (λ t ) is: m i (λ t ) is the mean function, which is set to 0, represents a diagonal matrix, and the diagonal elements are the noise variances of the t-th segment of the i-th spectrum at each wavelength, the elements in K i (λ t , λ t ) are calculated using the following formula:

[0020]

[0021] λ a represents the a-th element in λ t , λ b represents the b-th element in λ t , λ t is the wavelength of the t-th segment of the i-th spectrum after sampling, h and p are hyperparameters, and the maximum likelihood estimation algorithm is used to calculate the optimal h and p for each segment of the spectrum, and the log-likelihood function is as follows:

[0022]

[0023] wherein, D represents the number of elements in λ t , and the optimal hyperparameters are obtained by maximizing the log-likelihood function;

[0024] Given a new wavelength λ * , the t-th segment of the i-th spectrum at λ* The mean value of the flow rate μ i (λ * ), and the covariance Cov between each flow value. i (λ * ,λ * ):

[0025]

[0026]

[0027] Among them, f i (λ t Let λ be the flow rate value after sampling the t-th segment of the normalized i-th spectrum. * =λ t Calculate the spectral range of each segment at λ. t The mean of the flow rate at each location and the covariance between the flow rates.

[0028] In step S3, the weight of the p-th partial spectrum of the t-th segment of the i-th spectrum. The calculation formula is:

[0029]

[0030] n represents the number of absorption lines in this part of the spectrum, and z is the threshold set for assigning absorption lines from the original spectrum to the part of the spectrum.

[0031] The calculation process for step S4 is as follows:

[0032] Select the p-th portion of the t-th segment of the i-th and j-th spectra, and process the p-th portion of the t-th segment of the j-th spectrum as shown in the following formula:

[0033]

[0034] In the expression, each element represents the wavelength of the p-th partial spectrum within the t-th segment of the j-th spectrum, v represents velocity, and c represents the speed of light. The maximum likelihood estimation algorithm is used to calculate the radial velocity difference between the two partial spectra.

[0035]

[0036]

[0037] in, Indicates the wavelength of the two parts of the spectrum. This represents observational data for two parts of the spectrum. For a diagonal matrix, the diagonal elements are the noise variance of the two partial spectra at each wavelength, N represents the number of elements in the diagonal matrix.

[0038] The radial velocity difference between the pth partial spectrum of the tth segment of the ith spectrum and the pth partial spectrum of the tth segment of the jth spectrum is calculated according to the mean value of the flow values at λ t and the covariance between the respective flow values, and the simulated spectral data M is generated instead of the actual observation data in the above calculation process, and the multiple radial velocity differences ΔRV are calculated. The radial velocity difference between the pth partial spectrum of the tth segment of the ith spectrum and the pth partial spectrum of the tth segment of the jth spectrum is calculated according to the mean value of the flow values at λ and the covariance between the respective flow values, and the simulated spectral data M is generated instead of the actual observation data in the above calculation process, and the multiple radial velocity differences ΔRV

[0039]

[0040]

[0041] represent the radial velocity differences calculated by using the actual observation data.

[0042] In step S5, the radial velocity difference ΔRV ij(t) between the tth segment of the ith spectrum and the tth segment of the jth spectrum and the variance thereof are calculated according to the following formula:

[0043]

[0044]

[0045] is the weight of the pth partial spectrum of the tth segment of the ith spectrum.

[0046] In step S6, the radial velocity difference ΔRV ij between the ith spectrum and the jth spectrum is calculated according to the following formula:

[0047]

[0048] In step S7, the TV i of the ith spectrum is calculated according to the following formula: ΔRV ij is the radial velocity difference between the ith spectrum and the jth spectrum, and S represents the total number of spectra.

[0049] The application also provides a terminal device comprising a memory, a processor and a computer program stored in the memory; the processor executes the computer program to realize the steps of the above-mentioned method.

[0050] The application further provides a computer readable storage medium, which stores computer programs / instructions; the computer programs / instructions are executed by a processor to realize the steps of the above method.

[0051] Compared with the prior art, the application has the beneficial effects that:

[0052] (1) The application adopts a Gaussian process regression algorithm to process spectral data, reduces the error of the spectral data, and improves the calculation accuracy of the RV. The flow value at a certain wavelength in the spectrum is not completely independent of the flow values around it. According to the flow values around a wavelength, the error of the flow value at the point can be reduced, and more accurate spectral data can be obtained. The Gaussian process regression algorithm is adopted to process the spectral data, the correlation between the flow values is utilized, the error of the spectral data is reduced, more accurate flow values are obtained, and more accurate radial velocity differences of the star at different times can be obtained.

[0053] (2) The application calculates the RV of the star based on the periodicity of the radial velocity change of the star, and improves the calculation accuracy of the RV. Based on the periodicity of the radial velocity change, the radial velocity difference of the star at different times is calculated, and then the RV of the star is calculated according to the multiple radial velocity differences, so that the error of the RV is greatly reduced. Moreover, the more the number of star spectra used, the smaller the error of the RV. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 The figure is a flowchart of the star radial velocity calculation method of the embodiment 1 of the application.

[0055] Figure 2 The figure is a comparison diagram of the radial velocities calculated by the embodiment 1 of the application and the CCF algorithm. DETAILED DESCRIPTION

[0056] To make the purpose, technical scheme and advantages of the embodiments of the application clearer, the technical scheme of the embodiments of the application will be clearly and completely described below with reference to the drawings of the embodiments of the application. Obviously, the described embodiments are some of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments of the application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the application.

[0057] In this document, the terms "first," "second," and other similar words are not intended to imply any order, quantity, or importance, but are merely used to distinguish different elements. The terms "one," "a," and other similar words are not intended to indicate the existence of only one of the stated things, but rather that the description pertains to only one of the two stated things, which may include one or more. The terms "comprising," "including," and other similar words are intended to indicate a logical relationship, not a spatial relationship. For example, "A includes B" means that logically B belongs to A, not that spatially B is located inside A. Furthermore, the meanings of the terms "comprising," "including," and other similar words should be considered open-ended, not closed. For example, "A includes B" means that B belongs to A, but B does not necessarily constitute all of A; A may also include other elements such as C, D, and E.

[0058] Example 1

[0059] This embodiment uses 700 spectral data points from HD85512 as an example to illustrate the process in detail. Figure 1 The specific implementation steps.

[0060] Step 1: Normalize the spectral data and perform sampling. Each spectrum has a wavelength range of 380–690 nm. Divide the spectrum into 72 segments based on wavelength, discarding the extreme ends and retaining only the middle 50 segments for calculation. Normalize each segment using the following formula:

[0061]

[0062] In the formula f′ i (λ′ t f' represents the flow rate value of the i-th spectrum in segment t. i (λ′ t ) min Represents f′ i (λ′ t The minimum value in ), f′ i (λ′ t ) max Represents f′ i (λ′ t The maximum value in ) is used to process the error estimate of the spectral data. The formula used is:

[0063]

[0064] In the formula σ′ i (λ′ t ) represents f′ i (λ′ t) and error estimate of the i-th spectrum. The wavelength, flow rate value and error estimate of the i-th spectrum are denoted as λ t , f i (λ t ) and σ i (λ t ) respectively. The number of sample points is reduced to 50% of the original data by sampling, which effectively reduces the computational load of subsequent calculations.

[0065] Step 2: Construct a Gaussian process regression model for each segment of the preprocessed spectral data. Each segment of the spectrum includes multiple flow rate values, which follow a multivariate Gaussian distribution:

[0066]

[0067] m i (λ t ) represents the mean function, which is set to 0, is a diagonal matrix, and the diagonal elements are the noise variances of the t-th segment of the i-th spectrum at each wavelength. The elements in K i (λ t , λ t ) are calculated by a kernel function. For the spectral data of HD85512, the kernel function used is:

[0068]

[0069] The a-th element in λ t is denoted as λ a , and the b-th element in λ t is denoted as λ b . The element value in K i (λ t , λ t ) is calculated using k(λ a , λ b ). h and p are hyperparameters, and h and p are not the same for each segment of the spectrum. For each segment of the spectrum, the optimal h and p are calculated by maximizing the log-likelihood function, which is as follows:

[0070]

[0071] where, D represents the number of elements in λ t .

[0072] Let λ * = λ t , and calculate the mean μ i (λt ), and the covariance Cov between each flow value. i (λ t ,λ t ):

[0073]

[0074]

[0075] Step 3: Extract absorption lines from the spectrum and divide them into several groups based on their relative depth. Calculate the weight of each group relative to the original spectrum. Use a Gaussian process regression algorithm to establish a spectral model and obtain the values ​​of each spectral segment at λ. t The mean value of the flow rate μ i (λ t According to μ i (λ t Extract absorption lines from each spectral segment. Absorption lines with a relative depth greater than 0.1 and less than or equal to 0.35 constitute the first partial spectrum; absorption lines with a relative depth greater than 0.35 and less than or equal to 0.6 constitute the second partial spectrum; absorption lines with a relative depth greater than 0.6 and less than or equal to 0.8 constitute the third partial spectrum; and absorption lines with a relative depth greater than 0.8 constitute the fourth partial spectrum. Through the above operations, four partial spectra are extracted from each spectral segment. The weight of each partial spectrum relative to the original spectrum is calculated according to the following formula:

[0076]

[0077] n represents the number of absorption lines in each part of the spectrum, and z represents the lower boundary of the relative depth of the absorption lines in each part of the spectrum. For the first part of the spectrum, z = 0.1; for the second part of the spectrum, z = 0.35; for the third part of the spectrum, z = 0.6; and for the fourth part of the spectrum, z = 0.8.

[0078] Step 4: Calculate the line-of-sight velocity difference and variance between partial spectra extracted from different spectra using the same threshold. After Step 3, each spectrum segment is divided into multiple partial spectra, and the weight of each partial spectrum relative to the original spectrum is calculated. The p-th partial spectrum extracted from the t-th segment of the i-th and j-th spectra using the same threshold is selected, and the p-th partial spectrum of the t-th segment of the j-th spectrum is processed as follows:

[0079]

[0080] In the expression, the element represents the wavelength of the p-th partial spectrum in the t-th segment of the j-th spectrum, v represents velocity, and c represents the speed of light. The actual observed data of the two partial spectra extracted from the t-th segment of the i-th spectrum and the t-th segment of the j-th spectrum using the same threshold are denoted as […]. The noise variance is denoted as Wavelength is denoted as The line-of-sight velocity difference between the two partial spectra is calculated using the maximum likelihood estimation algorithm:

[0081]

[0082]

[0083] in, Let N be the number of elements in the array. Each pair of wavelength points in the equation is calculated using formula (4). The element values ​​in the matrix, and the selection of hyperparameters for the kernel function, utilize the optimal hyperparameter h obtained from the t-th segment of the i-th spectrum. it and ρ it .

[0084] Using the t-th segment of the i-th and j-th spectra respectively in λ t The mean of the flow rate values ​​and the covariance between each flow rate value are used to generate 30 pairs of simulated spectral data. Each time, unused simulated spectral data is selected from the simulated spectral data and replaced with the actual observed data in the above calculation process to recalculate the line-of-sight velocity difference. After obtaining multiple calculation results, their mean and variance are calculated to obtain the line-of-sight velocity difference between the partial spectra extracted from the two spectra using the same threshold. and its variance:

[0085]

[0086]

[0087] This represents the line-of-sight velocity difference calculated using simulated spectral data in the m-th experiment. This represents the line-of-sight velocity difference calculated using measured data.

[0088] Step 5: In Step 3, four partial spectra were extracted from an observed spectrum, and the weight of each partial spectrum relative to the original spectrum was calculated. Based on Step 4, the line-of-sight velocity difference and its variance between multiple pairs of partial spectra extracted using the same threshold were calculated. The multiple line-of-sight velocity differences were weighted and summed according to the weight values ​​obtained in Step 3 to obtain the line-of-sight velocity difference between the t-th segment of the i-th and j-th spectra, and their respective variances, as shown in the following formula:

[0089]

[0090]

[0091] Step 6: Calculate the radial velocity difference between different spectra. In step 1, 50 segments are selected from each spectrum for calculation. For the i-th and j-th spectra, 50 radial velocity differences and their variances are calculated according to the above process. In order to obtain accurate results, the radial velocity differences are weighted and summed according to the variances, so as to obtain the radial velocity difference between the i-th spectrum and the j-th spectrum:

[0092]

[0093] According to the above process, the radial velocity differences between different spectra are calculated.

[0094] Step 7: Calculate the RV of the star. According to the radial velocity difference between the i-th spectrum and other spectra, the RV of the star at different times is obtained:

[0095]

[0096] When i = j, ΔRV ij is regarded as 0. Through the above algorithm, the RV of the star at different times can be obtained.

[0097] In order to prove the superiority of the algorithm, the method of the embodiment is compared with the CCF algorithm, as shown in Figure 2 Table 1 is the root mean square value of the RV calculated by the two algorithms. From Table 1, it can be seen that the root mean square value of the RV calculated according to the method of the embodiment is smaller than the root mean square value of the RV calculated by the CCF, which means that the radial velocity calculation method disclosed in the embodiment can reduce the radial velocity calculation error caused by other factors and obtain more accurate RV.

[0098] Table 1 Root mean square value of RV

[0099]

[0100] Embodiment 2

[0101] Embodiment 2 of the present application provides a terminal device corresponding to the above-mentioned embodiment 1. The terminal device can be a processing device for a client, such as a mobile phone, a notebook computer, a tablet computer, a desktop computer, etc., to execute the method of the above-mentioned embodiment.

[0102] The terminal device of the embodiment comprises a memory, a processor and a computer program stored in the memory; the processor executes the computer program in the memory to realize the steps of the method of embodiment 1.

[0103] In some implementations, the memory can be a high-speed random access memory (RAM), and can also include a non-volatile memory, such as at least one disk memory.

[0104] In some implementations, the processor can be a central processing unit (CPU), a digital signal processor (DSP), or other types of general purpose processors, without limitation.

[0105] Embodiment 3

[0106] Embodiment 3 of the present application provides a computer readable storage medium corresponding to the above-mentioned embodiment 1, which stores computer programs / instructions. The computer programs / instructions are executed by the processor to implement the steps of the method of embodiment 1.

[0107] The computer readable storage medium can be a tangible device that maintains and stores instructions for use by an instruction execution device. The computer readable storage medium can be, for example but not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination thereof.

[0108] Those skilled in the art will understand that the embodiments of the present application can be provided as a method, system, or computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer-usable program code. The solutions in the embodiments of the present application can be implemented in various computer languages, such as object-oriented programming languages Java and interpreted scripting language JavaScript, etc.

[0109] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems), and computer program products of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The device that implements the functions specified in one flow or multiple flows and / or blocks Figure 1 The device that implements the functions specified in one flow or multiple flows and / or blocks

[0110] These computer program instructions can also be loaded into computer or other programmable data processing devices, so that a series of operations steps are performed on the computer or other programmable devices to generate computer-implemented processes, thus the instructions executed on the computer or other programmable devices provide the function of realizing the processes specified in the flowcharts Figure 1 one flowchart or multiple flowcharts and / or blocks Figure 1 one block or multiple blocks.

[0111] Although the preferred embodiments of the application have been described, those skilled in the art will be able to make additional modifications and variations to these embodiments without departing from the spirit and scope of the application. Accordingly, it is intended that the appended claims be construed to include all such modifications and variations as fall within the scope of the application.

[0112] Obviously, various modifications and changes are possible in the present application without departing from the spirit and scope of the application. It is therefore intended that the present application encompass all such modifications and changes as fall within the scope of the claims and their equivalents.

Claims

1. A method of calculating stellar radial velocity, characterized by, The method comprises the following steps: S1, acquiring S spectra, dividing each spectrum into T segments, and pre-processing each segment of spectrum data; S2, constructing a Gaussian process regression model for each segment of spectrum using the pre-processed spectrum data, and calculating the mean of the flow values at different wavelengths and the covariance between the flow values for each segment of spectrum; S3, extracting absorption lines in each segment of spectrum according to the mean of the flow values at different wavelengths, grouping absorption lines with a relative depth in the same range into a group to form a partial spectrum, and calculating the weight of each partial spectrum relative to the original spectrum; S4, calculating the radial velocity difference and variance between partial spectra obtained by segmenting different spectra with the same threshold value; S5, repeating step S4, calculating a plurality of radial velocity differences and variances using a plurality of partial spectra of the same segment of two spectra, and calculating the radial velocity difference and variance between the same segments of the two spectra by weighted summation of the plurality of radial velocity differences using the weight described in step S3; S6, repeating steps S4-S5, calculating a plurality of radial velocity differences and variances according to a plurality of segment data of two spectra, and calculating the radial velocity difference between the two spectra; S7, obtaining the radial velocity of any spectrum according to the radial velocity difference between the spectrum and other spectra.

2. The method of claim 1, wherein, The specific implementation process of step S1 comprises: Compute f i (λ′ t ) where λ′ t is the wavelength of the tth segment of the ith spectrum, f′ i (λ′ t ) is the flux value of the tth segment of the ith spectrum, the minimum value in f′ i (λ′ t ) is denoted f′ i (λ′ t ) min , and the maximum value is denoted f′ i (λ′ t ) max , and σ′ i (λ′ t ) is the error estimate for each flux value; f i (λ′ t ) is the normalized flux value of the tth segment of the ith spectrum, and σ i (λ′ t ) is the error estimate for f i (λ′ t ). The t-th segment of the normalized i-th spectrum is sampled, and the sampled wavelength is denoted as λ t , the flow value is f i (λ t ), and the error estimation value is σ i (λ t ).

3. The method of claim 1, wherein, In step S2, the construction process of the Gaussian process regression model for the tth segment of the ith spectrum comprises: The observed spectrum is represented as: f i (λ t ) = y i (λ t ) + n i (λ t ), n i (λ t ) represents Gaussian noise, y i (λ t ) ~ GP(m i (λ t ), K i (λ t , λ t )), GP(m i (λ t ), K i (λ t , λ t )) represents a Gaussian process, m i (λ t ) is a mean function, set to 0, K i (λ t , λ t ) is a covariance function; the prior distribution of f i (λ t ) is: represents a diagonal matrix, the diagonal elements are the noise variances of the tth segment of the ith spectrum at each wavelength, the elements in K i (λ t , λ t ) are calculated using the following formula: λ a represents the a-th element in λ t represents the b-th element in λ b represents the a-th element in λ t represents the b-th element in λ t is the spectral wavelength after sampling the t-th segment of the i-th normalized spectrum, and h and p are hyperparameters; Given a new wavelength λ * , the mean μ * (λ i ) and the covariance Cov * (λ i ,λ * ) between the individual flow values are calculated for the t-th segment of the i-th spectrum at λ * using the following equations: where f i (λ t ) is the flow value of the tth segment of the ith normalized spectrum after sampling; let λ * = λ t , the mean value of the flow values of each segment of the spectrum at λ t and the covariance between each flow value are calculated, respectively.

4. The method of claim 1, wherein, The weight of the pth partial spectrum in the tth segment of the ith spectrum in step S3 The calculation formula is: n represents the number of absorption lines in the partial spectrum, and z is a threshold value set for distributing the absorption lines in the original spectrum to the partial spectrum.

5. The method of claim 1, wherein, The calculation process of step S4 comprises: The pth partial spectrum of the tth section of the ith spectrum is selected, and the pth partial spectrum of the tth section of the jth spectrum is processed: The element in the formula represents the wavelength of the pth partial spectrum of the tth section of the jth spectrum, v represents the velocity, and c represents the speed of light. The calculation formula of the radial velocity difference between the pth partial spectrum of the tth segment of the ith spectrum and the pth partial spectrum of the tth segment of the jth spectrum is: wherein, denotes the wavelengths of the two partial spectra, denotes the observation data of the two partial spectra, denotes the observation data as the likelihood corresponding to the parameter v, denotes a diagonal matrix with diagonal elements being the noise variances of the two partial spectra at each wavelength, N denotes the number of elements in . The mean of the flow values at λ t The mean of the flow values at λ The radial velocity difference between the pth partial spectrum of the tth segment of the ith spectrum and the pth partial spectrum of the tth segment of the jth spectrum is calculated according to M is the total number of simulation, is the observed radial velocity difference.

6. The method of claim 1, wherein, In step S6, the calculation formula of the radial velocity difference between the ith spectrum and the jth spectrum is: wherein, is the weight of the p-th partial spectrum of the t-th segment of the i-th spectrum.

7. The method of claim 1, wherein, In step S7, the radial velocity RV of the i-th spectrum i The calculation formula is: ΔRV ij is the radial velocity difference between the i-th spectrum and the j-th spectrum, and S represents the total number of spectra.

8. A terminal device comprising a memory, a processor, and a computer program stored on the memory; characterized in that, The processor executes the computer program to implement the steps of the method of any one of claims 1-7.

9. A computer readable storage medium having stored thereon computer programs / instructions; characterized in that, The computer program / instructions are executed by the processor to implement the steps of the method of any one of claims 1-7.

Images