Off-line data searching method and system applied to Fourier infrared spectrum analyzer

Abstract

The invention provides an off-line data searching method and system applied to a Fourier infrared spectrum analyzer, relates to the field of digital signal processing, and aims to solve the problems that the effect of data acquired by a Fourier spectrometer is poor, the influence of a noise change trend term is not considered, and the data is not easy to search. And the noise data which does not contain the spectrogram is misjudged as the interferogram data. The method comprises the following steps: acquiring signal data containing an interference pattern and noise; performing coarse search and fine search on the signal data containing the interferogram and the noise to obtain the interferogram of the signal data; according to the method, the coarse search and the fine search are combined, so that the calculation complexity of the search is reduced, and the accuracy of the search is improved; interferogram signals and noise are detected and distinguished by using a method of combining a polynomial fitting trend term removal algorithm and an autocorrelation algorithm.

Description

Technical Field

[0001] The invention belongs to the field of digital signal processing, and in particular relates to an off-line data search method and system applied to a Fourier infrared spectrometer. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] The Fourier spectrometer uses the optical path difference produced by the movement of the moving mirror of the Michelson interferometer to perform time modulation on the incident light beam to obtain an interference pattern formed by various spectral components, and obtains spectral information using the inverse Fourier transform. It has the advantages of fast wavelength scanning speed, good resolution, high accuracy and multiple channels, so it has been widely studied and applied in academic and industrial fields. With the development of related electronic, mechanical, control and optical technologies and the maturity of market conditions, more and more research institutions have invested in the research and development and application of Fourier spectrometers.

[0004] During the development of the rotating mirror Fourier spectrometer, a mercury cadmium telluride detector established a corresponding relationship between the detection signal and the incident radiation, so that the interference signal of the incident light can be obtained. The digital acquisition card samples the interference signal to obtain a digital signal. The moving mirror of the spectrometer rotates one circle to form a bilateral interference pattern. When the moving mirror rotates n circles, a digital signal containing n bilateral interference patterns will be collected. Due to the limitation of the maximum working angle of the lens of the rotating mirror Fourier spectrometer, the interference pattern data is not continuously generated. In an experiment, there are often tens of millions of digital sampling points. Especially when the signal-to-noise ratio is low, it is difficult to find the interference pattern directly in the data. Therefore, when processing the interference signal offline, the first problem faced is to search for the interference pattern data from a large amount of data, and only on this basis can the interference pattern be subsequently filtered, spectral inversion and other processing be performed.

[0005] The inventors found that the existing methods generally directly use the traditional autocorrelation algorithm to search for interference pattern data, but the actual measured data of this method is poor and the interference pattern cannot be accurately and effectively searched out. This is because the signal-to-noise ratio is low, and the noise data generally has a certain change trend. The method of directly using the main and secondary peak ratios of the autocorrelation function for data screening does not consider the influence of the noise change trend term, resulting in the noise data that does not contain the spectrum being misjudged as interference pattern data, and then the search effect for the data collected by the Fourier spectrometer is poor, and the interference pattern data cannot be accurately and effectively searched out. Summary of the invention

[0006] In order to overcome the deficiencies of the above-mentioned prior art, the present invention provides an offline data search method and system applied to a Fourier transform infrared spectrometer, which adopts a combination of coarse search and fine search to screen out a bilateral interference pattern centered on a maximum value from a large amount of data, thereby reducing the computational complexity of the search and improving the accuracy of the search.

[0007] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0008] A first aspect of the present invention provides an offline data search method applied to a Fourier transform infrared spectrometer, comprising:

[0009] Acquire signal data containing interference pattern and noise;

[0010] Performing a rough search on the signal data containing the interference pattern and the noise to obtain a data segment after the rough search;

[0011] Perform fine search on the data segment after the rough search to obtain the interference pattern of the signal data;

[0012] In both the coarse search and the fine search, polynomial fitting is used to remove the trend term, and then the autocorrelation algorithm is used to detect and distinguish the interference pattern and the noise.

[0013] As an implementation method, a rough search is performed on signal data containing interference patterns and noise to obtain data segments after the rough search. The specific process is as follows:

[0014] Segmenting the signal data containing the interference pattern and the noise to obtain a number of segmented data;

[0015] Calculate the mean of the segmented data;

[0016] Subtract the mean from each data of the segmented data to obtain the segmented data with the mean subtracted;

[0017] Calculate the maximum value of the segmented data minus the mean;

[0018] Divide the segmented data minus the mean by the maximum value of the segmented data minus the mean, and normalize the division result;

[0019] Use polynomial fitting to calculate the trend term of segmented data;

[0020] Remove the trend item of segmented data;

[0021] Perform autocorrelation on the segmented data without trend items to obtain autocorrelation data;

[0022] Normalizing the autocorrelation data to obtain normalized autocorrelation data;

[0023] According to the normalized autocorrelation data, the sub-correlation peak value of the autocorrelation data is calculated;

[0024] If the secondary correlation peak value is greater than or equal to the first threshold, the segmented data is a data segment after a rough search.

[0025] As an implementation mode, after obtaining the data segment obtained by the rough search, the data segment obtained by the rough search is saved, specifically:

[0026] The segment number of each data segment obtained by the rough search is stored in the segment number set;

[0027] Calculate the maximum value of the segmented data and save the number of the maximum value in the index collection.

[0028] As an implementation method, when performing a fine search on a data segment after a rough search, the data segment after the rough search is first corrected, and the specific process is as follows:

[0029] Traverse the segment number set and index set in sequence;

[0030] According to the segment number and the maximum value number, determine the maximum value of the entire data;

[0031] A data segment is intercepted with the maximum value as the center, and the intercepted data segment is corrected to obtain a corrected data segment.

[0032] As an implementation method, after the data segment after the rough search is corrected, the detailed search is continued as follows:

[0033] Calculate the mean of the corrected data segment;

[0034] Subtract the mean from the corrected data segment to obtain a mean-subtracted data segment;

[0035] Calculate the maximum value of the data segment minus the mean;

[0036] Divide the data segment with the mean subtracted by the maximum value of the data segment with the mean subtracted, and normalize the division result;

[0037] Calculate the trend term of the normalized data segment;

[0038] Remove the trend item of the normalized data segment;

[0039] Perform autocorrelation on the data segment without the trend term to obtain autocorrelation data;

[0040] Normalizing the autocorrelation data to obtain normalized autocorrelation data;

[0041] According to the normalized autocorrelation data, the sub-correlation peak value of the autocorrelation data is calculated;

[0042] If the secondary correlation peak is less than the second threshold, the data segment is a noise data segment and is removed.

[0043] If the secondary correlation peak value is greater than or equal to the second threshold, the data segment is retained, and an interference pattern with the maximum value as the symmetry center is searched therefrom.

[0044] As an implementation mode, if the reserved data segment includes a noise data segment, the reserved data segment is continuously finely searched until an interference pattern is found.

[0045] A second aspect of the present invention provides an offline data search system for a Fourier transform infrared spectrometer, comprising:

[0046] A data acquisition module, used for acquiring signal data containing interference patterns and noise;

[0047] A coarse search module is used to perform a coarse search on signal data containing interference patterns and noise to obtain a data segment after the coarse search;

[0048] A fine search module is used to perform a fine search on the data segment after the rough search to obtain an interference pattern of the signal data;

[0049] In both the coarse search and the fine search, polynomial fitting is used to remove the trend term, and then the autocorrelation algorithm is used to detect and distinguish the interference pattern and the noise.

[0050] A third aspect of the present invention provides a computer device, comprising a memory, a processor, and a program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the method described in the first aspect of the present invention are implemented.

[0051] The fourth aspect of the present invention aims to provide a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps in the method described in the first aspect of the present invention.

[0052] A fifth aspect of the present invention aims to provide a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute the method and functions described in the first aspect of the present invention.

[0053] One or more of the above technical solutions have the following beneficial effects:

[0054] In this embodiment, the offline data search method applied to the Fourier spectrometer adopts a combination of coarse search and fine search, which can screen out the bilateral interference pattern centered on the maximum value from a large amount of data, thereby reducing the computational complexity of the search and improving the accuracy of the search.

[0055] In this embodiment, according to the different effects of trend terms on noise data and interference pattern data, a method combining polynomial fitting to remove trend terms and an autocorrelation algorithm is used to perform data search, thereby realizing detection and distinction of interference pattern signals and noise.

[0056] In this embodiment, different fitting orders and secondary correlation peak thresholds are set according to the differences between the coarse search and fine search stages, thereby reducing the computational complexity.

[0057] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0059] Figure 1 This is a flow chart of a discrete data retrieval method applied to a Fourier infrared spectrometer in the first embodiment;

[0060] Figure 2 This is a time domain diagram of the collected data of the spectrometer in the first embodiment. DETAILED DESCRIPTION

[0061] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0062] It should be noted that the terms used herein are for describing specific embodiments only and are not intended to be limiting of exemplary embodiments according to the present invention.

[0063] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0064] Embodiment 1

[0065] This embodiment discloses an offline data search method applied to a Fourier infrared spectrometer.

[0066] In order to more clearly illustrate the present embodiment, an implementation process of an offline data search method applied to a Fourier transform infrared spectrometer can be specifically described as follows:

[0067] An off-line data search method applied to a Fourier transform infrared spectrometer comprises:

[0068] S1, obtaining signal data containing interference pattern and noise;

[0069] S2, performing a rough search on the signal data containing the interference pattern and the noise to obtain a data segment after the rough search;

[0070] S3, performing a fine search on the data segment after the rough search to obtain an interference pattern of the signal data;

[0071] like Figure 1 As shown, in step S1, signal data containing interference pattern and noise is acquired.

[0072] In this embodiment, when the moving mirror of the rotating mirror Fourier spectrometer rotates one circle, that is, 360°, the maximum working angle of the moving mirror is 15°, and the two beams generated by the beam splitter only interfere in the angle range of -15° to 15°. In other words, when the moving mirror rotates once, in the collected data, only one section of data is interference signal data, and the rest is noise data. Figure 2 As shown in the figure, the rotating mirror rotates for many circles continuously, and multiple signal interference occurs, and 36.7 million data points are collected. Due to the low signal-to-noise ratio and the large amount of data, the interference pattern is difficult to distinguish by naked eyes. The small tip in the circle is a section of data that interferes, and the details of the interference pattern can be seen through the enlarged image pointed by the arrow. A data search method in this embodiment is to search for interference pattern data that interferes multiple times from 30 million data points.

[0073] In order to reduce the computational complexity of the search, this embodiment adopts a method combining coarse search and fine search. First, the interference data is roughly searched out by relaxing the search threshold, and then the search accuracy is further improved from the searched data segment by increasing the search threshold.

[0074] like Figure 1 As shown, in step S2, a coarse search is performed on the signal data containing the interference pattern and the noise to obtain a data segment after the coarse search.

[0075] In this embodiment, a coarse search is performed on the signal data containing the interference pattern and the noise to obtain the data segment after the coarse search. The specific process is as follows:

[0076] S2-1. Segment the signal data containing the interference pattern and the noise to obtain a plurality of segmented data.

[0077] When an interference occurs, the length of the interference pattern data generated is very short relative to the total data length. Therefore, the data needs to be processed in segments and the interference pattern data needs to be searched segment by segment. This can improve the efficiency and accuracy of subsequent searches.

[0078] In this embodiment, the data Data of more than 30 million points are divided into M segments, and the number of data points in each segment is N.

[0079] S2-2. Calculate the mean of each segmented data.

[0080] In this embodiment, the mth segment data is D m , the data point can be represented as {D m [1], ... D m [N]}. Calculate the mean of the mth segment of data, add up each data in the mth segment and divide by the number of data points N to get the mean V of the mth segment of data.

[0081] S2-3, subtract the mean from each data of the segmented data to obtain the segmented data with the mean subtracted;

[0082] In this embodiment, in order to remove the influence of the DC component on data processing, the mth segment data D m Subtract the mean, the formula is:

[0083] D m ={D m [1] -V, ... D m [N]-V} (1)

[0084] Among them, D m [1]-V represents the first data element of the mth segment minus the mean value V, D m [N]-V represents the Nth data element in the mth segment minus the mean value V.

[0085] S2-4. Calculate the maximum value of the segmented data minus the mean value.

[0086] In this embodiment, the mth segment data D minus the mean is calculated. m The maximum value of is:

[0087] val=max[abs(D m )] (2)

[0088] Among them, val represents D minus the mean m The maximum value of the segment data, max[] means to take the maximum value of the segment data, abs() means to m The segment data takes the absolute value.

[0089] S2-5. Divide the segmented data from which the mean is subtracted by the maximum value of the segmented data from which the mean is subtracted, and normalize the division result.

[0090] In this embodiment, the mth segment data D m Each data D m [i] are normalized by dividing by the maximum value, and the formula is:

[0091] D m ′[i]=D m[i] / val, i=1,...N (3)

[0092] Among them, D m ′[i] is the normalized data element of the mth segment, and val represents D m The maximum value of the segment data.

[0093] S2-6. Calculate the trend item of the normalized segmented data.

[0094] In this embodiment, the calculation of the mth segment D m Each data D m The trend term X of ′[i] m , for segment m D m Each data D m ′[i]Using the polynomial fitting method, we get N fitting data points, which is the mth segment D m Each data D m The trend term X of ′[i] m .

[0095] In this embodiment, in order to reduce the computational complexity, the fitting order is set to 3.

[0096] S2-7. Remove the trend item of the segmented data.

[0097] Since the data segment containing only noise will have a strong trend term, resulting in a large autocorrelation coefficient, once the trend term is removed, the data change trend is flattened and the correlation coefficient decreases rapidly. However, after the trend term is removed from the interference pattern data, its correlation coefficient does not change much. In order to search for the interference pattern data, the method of removing the trend term of the segmented data is adopted.

[0098] In this embodiment, the mth segment D m Each data D m The trend term in ′[i] is removed. The data formula after removing the trend term is:

[0099]

[0100] in, Indicates the mth segment of data D m The i-th data element after removing the trend term, X m [i] represents the trend item of the i-th data element, and N represents the number of data points in the data segment.

[0101] S2-8. Perform autocorrelation on the segmented data with the trend item removed to obtain autocorrelation data.

[0102] In this embodiment, after removing the trend term, D m Do autocorrelation on the segment data to get the autocorrelation data R m.

[0103] S2-9. Normalize the autocorrelation data to obtain normalized autocorrelation data.

[0104] In this embodiment, R m Normalize and get the normalized autocorrelation data. The formula is:

[0105] R m [i] = R m [i] / R m [N], i = 1, 2, ... 2N-1 (5)

[0106] Among them, R m [i] represents the autocorrelation data R m The i-th element, R m [N] stands for R m The Nth element of .

[0107] After normalization of the autocorrelation sequence, R m [N] is the maximum value and is equal to 1.

[0108] S2-10. Calculate the secondary correlation peak value of the autocorrelation data based on the normalized autocorrelation data.

[0109] In this embodiment, let R m [N]=0, find the sequence abs(R m ), and take the maximum value at this time as the secondary correlation peak, assuming that the secondary correlation peak is Peak m .

[0110] Generally speaking, the sub-correlation peak of a data segment containing only noise is small, while the sub-correlation peak of a segment containing interference pattern data is large. Interference pattern data is selected based on the size of the sub-correlation peak.

[0111] S2-11. If the secondary correlation peak value is greater than or equal to the first threshold, the segmented data is a data segment after a rough search.

[0112] In this embodiment, steps S2-1 to S2-11 are performed on M segments of data in sequence. If the secondary correlation peak Peak corresponding to the mth segment of data is m ≥0.3, it is considered that the mth segment contains interference pattern data. At this time, the number m is saved in the segment number set index, and D is calculated at the same time. m The maximum value of the data in is:

[0113]

[0114] The maximum value val′ is indexed as Pos mAnd save the maximum value number in the index set Pos.

[0115] After the above steps, polynomial fitting is introduced, mainly because the data segment containing only noise will have a strong trend term, resulting in a large autocorrelation coefficient. Once the trend term is removed, the data change trend is flattened and the correlation coefficient decreases rapidly. However, after the trend term is removed from the interference pattern data, the correlation coefficient does not change much.

[0116] like Figure 1 As shown, in step S3, a fine search is performed on the data segment after the coarse search to obtain an interference pattern of the signal data.

[0117] Because there is no prior information, the threshold in the coarse search is set relatively wide, and because the signal-to-noise ratio is low, the data segment corresponding to the number in the set index is likely to be noise data and is mistakenly searched and filtered out. Therefore, the second stage of the fine search process is required.

[0118] Perform fine search on the data segment after rough search to obtain the interference diagram of signal data. The specific process is as follows:

[0119] S3-1. Traverse the segment number set and index set in sequence.

[0120] In this embodiment, the elements in index are taken out one by one for calculation. At this time, the element taken out from index is i, and the corresponding element in Pos is Pos i , that is, the index Pos of the maximum value data in the i-th segment of data i .

[0121] For example, if the data segment traversed from the segment number set index is the 3rd segment, the number of the maximum value in the 3rd segment is Pos3.

[0122] S3-2. Determine the maximum value of the entire data according to the segment number and the maximum value number.

[0123] In this embodiment, according to the segment number i and the maximum value number Pos i , find the maximum value D of the i-th segment of data i [Pos i ], which is the maximum value of the entire data.

[0124] For example, the maximum value of the third segment of data traversed is D3[Pos3], that is, the maximum value of the entire data is D3[Pos3].

[0125] S3-3, intercepting a data segment with the maximum value as the center, and correcting the intercepted data segment to obtain a corrected data segment.

[0126] In this embodiment, in the entire data Data, Di [Pos i ] as the center, take N / 2 data before and after, and update data segment D i This is because the maximum value of the data segment screened out by the rough search is generally not located at the center of the data segment. Therefore, the data segment is re-cut with the maximum value as the center, and the cut data is corrected to ensure that the screened data is a bilateral interference pattern centered on the maximum value.

[0127] For example, taking D3[Pos3] as the center, take N / 2 data before and after, where N is the number of data points in the data segment. If N is 10000, then taking the maximum value D3[Pos3] of the third segment as the center, take 5000 data before and after, and the corrected data segment is obtained.

[0128] S3-4. Repeat steps S2-1 to S2-11 for the modified data segment.

[0129] Since the data segments being searched may contain data segments that contain only noise, a fine search is required to remove the noise data segments. The search process steps are the same as the coarse search, except that the parameter settings are changed.

[0130] The specific operation process is the same as that of steps S2-1 to S2-11, specifically:

[0131] S3-4-1. Calculate the mean of the corrected data segment.

[0132] S3-4-2. Subtract the mean from the corrected data segment to obtain the data segment minus the mean.

[0133] S3-4-3. Calculate the maximum value of the data segment minus the mean.

[0134] S3-4-4. Divide the data segment from which the mean is subtracted by the maximum value of the data segment from which the mean is subtracted, and normalize the division result.

[0135] S3-4-5. Calculate the trend item of the normalized data segment.

[0136] In this embodiment, the number of indexes in the segment number set index is relatively small, and the fitting order is set to 6 to ensure the fitting accuracy.

[0137] S3-4-6. Remove the trend item of the normalized data segment.

[0138] S3-4-7. Perform autocorrelation on the data segment without the trend item to obtain autocorrelation data.

[0139] S3-4-8. Normalize the autocorrelation data to obtain normalized autocorrelation data.

[0140] S3-4-9. Calculate the secondary correlation peak value of the autocorrelation data based on the normalized autocorrelation data.

[0141] S3-4-10. If the secondary correlation peak is less than the second threshold, the data segment is a noise data segment and is removed. If the secondary correlation peak is greater than or equal to the second threshold, the data segment is retained and an interference pattern with the maximum value as the symmetry center is searched therefrom.

[0142] In this embodiment, when setting the secondary correlation peak value Peak m ≥0.5, it is determined that the data segment contains an interference pattern data segment, and the interference pattern with the maximum value as the symmetry center is searched from the data segment. m <0.5: This data segment is a noise data segment and is removed.

[0143] After the above steps, the segment number set index is traversed to find the bilateral interference pattern with the maximum value as the symmetry center.

[0144] S3-4-11. If the retained data segment contains a noise data segment, the retained data segment is continuously searched until the interference pattern is found.

[0145] If there are still noise data segments in the filtered data segments at this time, repeat the detailed search.

[0146] This embodiment adopts a method combining coarse search and fine search to improve the accuracy of search; at the same time, different fitting orders and secondary correlation peak thresholds are set according to the different coarse search and fine search stages, thereby reducing the calculation complexity.

[0147] Embodiment 2

[0148] The purpose of this embodiment is to provide an offline data search system applied to a Fourier transform infrared spectrometer, comprising:

[0149] A data acquisition module, used for acquiring signal data containing interference patterns and noise;

[0150] A coarse search module is used to perform a coarse search on signal data containing interference patterns and noise to obtain a data segment after the coarse search;

[0151] A fine search module is used to perform a fine search on the data segment after the rough search to obtain an interference pattern of the signal data;

[0152] In both the coarse search and the fine search, polynomial fitting is used to remove the trend term, and then the autocorrelation algorithm is used to detect and distinguish the interference pattern and the noise.

[0153] Based on an offline data search system applied to a Fourier infrared spectrometer, the steps of an offline data search method applied to a Fourier infrared spectrometer in the first embodiment are implemented.

[0154] Embodiment 3

[0155] The purpose of this embodiment is to provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the program.

[0156] Embodiment 4

[0157] The purpose of this embodiment is to provide a computer-readable storage medium.

[0158] A computer-readable storage medium stores a computer program, which executes the steps of the above method when executed by a processor.

[0159] Embodiment 5

[0160] The purpose of this embodiment is to provide a computer program product containing instructions, which, when running on a computer, enables the computer to execute the methods and functions involved in any of the above embodiments.

[0161] The steps involved in the apparatus of the above embodiment correspond to the method embodiment 1, and the specific implementation method can refer to the relevant description part of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0162] Those skilled in the art should understand that the modules or steps of the present invention described above can be implemented by a general-purpose computer device, or alternatively, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0163] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. An off-line data search method applied to a Fourier transform infrared spectrometer, characterized in that: include: Acquire signal data containing interference pattern and noise; Performing a rough search on the signal data containing the interference pattern and the noise to obtain a data segment after the rough search; Perform fine search on the data segment after the rough search to obtain the interference pattern of the signal data; In both the coarse search and the fine search, polynomial fitting is used to remove the trend term, and then the autocorrelation algorithm is used to detect and distinguish the interference pattern and the noise.

2. The off-line data search method for Fourier transform infrared spectrometer according to claim 1, characterized in that: A rough search is performed on the signal data containing interference patterns and noise to obtain the data segment after the rough search. The specific process is as follows: Segmenting the signal data containing the interference pattern and the noise to obtain a number of segmented data; Calculate the mean of each segmented data; Subtract the mean from each data of the segmented data to obtain the segmented data with the mean subtracted; Calculate the maximum value of the segmented data minus the mean; Divide the segmented data minus the mean by the maximum value of the segmented data minus the mean, and normalize the division result; Use polynomial fitting to calculate the trend term of segmented data; Remove the trend item of segmented data; Perform autocorrelation on the segmented data without trend items to obtain autocorrelation data; Normalizing the autocorrelation data to obtain normalized autocorrelation data; According to the normalized autocorrelation data, the sub-correlation peak value of the autocorrelation data is calculated; If the secondary correlation peak value is greater than or equal to the first threshold, the segmented data is a data segment after a rough search.

3. The off-line data search method for Fourier transform infrared spectrometer according to claim 1, characterized in that: After obtaining the data segment after the rough search, the data segment after the rough search is saved, specifically: The segment number of each data segment after the rough search is stored in the segment number set; Calculate the maximum value of the segmented data and save the number of the maximum value in the index collection.

4. The off-line data search method for Fourier transform infrared spectrometer according to claim 1, characterized in that: When performing a fine search on a data segment after a rough search, the data segment after the rough search is first corrected. The specific process is as follows: Traverse the segment number set and index set in sequence; According to the segment number and the maximum value number, determine the maximum value of the entire data; A data segment is intercepted with the maximum value as the center, and the intercepted data segment is corrected to obtain a corrected data segment.

5. The off-line data search method for Fourier transform infrared spectrometer according to claim 1, characterized in that: After the data segment after the rough search is corrected, the detailed search is continued. The specific process is as follows: Calculate the mean of the corrected data segment; Subtract the mean from the corrected data segment to obtain a mean-subtracted data segment; Calculate the maximum value of the data segment minus the mean; Divide the data segment with the mean subtracted by the maximum value of the data segment with the mean subtracted, and normalize the division result; Calculate the trend term of the normalized data segment; Remove the trend item of the normalized data segment; Perform autocorrelation on the data segment without the trend term to obtain autocorrelation data; Normalizing the autocorrelation data to obtain normalized autocorrelation data; According to the normalized autocorrelation data, the sub-correlation peak value of the autocorrelation data is calculated; If the secondary correlation peak is less than the second threshold, the data segment is a noise data segment and is removed. If the secondary correlation peak is greater than or equal to the second threshold, the data segment is retained and an interference pattern with the maximum value as the symmetry center is searched therefrom.

6. The off-line data search method for Fourier transform infrared spectrometer according to claim 1, characterized in that: If the reserved data segment contains a noise data segment, the reserved data segment is continuously searched until an interference pattern is found.

7. An off-line data search system for a Fourier transform infrared spectrometer, characterized in that: include: A data acquisition module, used for acquiring signal data containing interference patterns and noise; A coarse search module is used to perform a coarse search on signal data containing interference patterns and noise to obtain a data segment after the coarse search; A fine search module is used to perform a fine search on the data segment after the rough search to obtain an interference pattern of the signal data; In both the coarse search and the fine search, polynomial fitting is used to remove the trend term, and then the autocorrelation algorithm is used to detect and distinguish the interference pattern and the noise.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of an offline data search method applied to a Fourier transform infrared spectrometer as described in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of an offline data search method applied to a Fourier infrared spectrometer as described in any one of claims 1 to 6 are performed.

10. A computer program product, when it is run on a computer, characterized in that The computer is enabled to execute the off-line data search method and function applied to the Fourier infrared spectrometer as described in any one of claims 1-6.

Images