CN-115859080-B - Optical fiber Bragg grating signal peak detection algorithm based on asymmetric Gaussian model

CN115859080BCN 115859080 BCN115859080 BCN 115859080BCN-115859080-B

Abstract

The invention discloses an asymmetric Gaussian model-based optical fiber Bragg grating signal peak detection algorithm, which comprises the steps of smoothing an acquired optical fiber Bragg grating spectrum signal by adopting a five-point sliding average filtering method to obtain a smoothed spectrum signal, conducting primary derivation on the smoothed spectrum signal, determining the size of a window according to the positions of the maximum value point and the minimum value point of a derivative, resampling the spectrum signal in the window, reducing the sampling rate, adopting Gaussian fitting on the resampled spectrum signal in the window, preliminarily determining the position of a peak value, introducing an asymmetric Gaussian function model, correcting the position of the peak value by combining the detected peak value with the asymmetric Gaussian model, and finishing accurate positioning of the spectrum peak value.

Inventors

LIU SONG
SUN ZHIPENG
DONG CHENG
QIU DA
QIAN KAI
CHEN SHIQIANG
ZHANG JIANQIANG
TIAN FANG
ZHANG TINGTING

Assignees

湖北民族大学
武汉国电武仪电气股份有限公司

Dates

Publication Date: 20260505
Application Date: 20221123

Claims (3)

1. The optical fiber Bragg grating signal peak value detection algorithm based on the asymmetric Gaussian model is characterized by comprising the following steps of: s1, smoothing the collected optical fiber Bragg grating spectrum signal by adopting a five-point sliding average filtering method to obtain a smoothed spectrum signal; s2, conducting primary derivation on the spectrum signal after the smoothing processing, and determining the size of a window according to the positions of the maximum value point and the minimum value point of the derivative; S3, resampling the spectrum signals in the window, reducing the sampling rate, adopting Gaussian fitting to the resampled spectrum signals in the window, and preliminarily determining the position of the peak value, wherein the method comprises the following steps of: S31, the fiber Bragg grating signal is expressed as: (2) Wherein: Is the wavelength of the fiber bragg grating spectrum, Is the spectral center wavelength of the fiber bragg grating, For a 3dB bandwidth, A is the amplitude of the reflectance spectrum; Taking the logarithm of two sides of the formula (2) at the same time to obtain: (3) And (3) making: , , , The formula (3) is simplified as: (4) the least square method is used to determine the values of a, b and c in equation (4), the center wavelength is: (5) s32, making the point B be the theoretical peak value And D is the maximum point calculated by the formula (5), and the maximum point is substituted into a function of second-order Gaussian fitting when the sampling rate is reduced: (6) Wherein a1, a2, b1, b2, c1 and c2 are parameters of a second-order Gaussian function, and D point coordinates are determined , ) To Selecting data intervals as criteria Two adjacent data points A and C of D are determined, and the coordinates are respectively [ ] , ),( , ) Bringing points A, D, and C into formula (4), respectively, yields: (7) Calculating the values of a, b and c, and bringing the 3 values into formula (5) (8) Calculating a center wavelength from the formula (8), wherein a peak value of the center wavelength is a preliminarily determined peak value position; s4, introducing an asymmetric Gaussian function model to correct the initially determined peak position, wherein the method specifically comprises the following steps: And performing verification compensation on the initially determined peak position, wherein the formula is as follows: (9) Asymmetric gaussian model: (10) Wherein, the The time point corresponding to the peak point is obtained for the gaussian fitting function, The number of time point samples for the left part of the gaussian fitting function, For the sampling number of the time points of the right part of the Gaussian fitting function, the judgment of the asymmetric Gaussian function is based on 2 second-order parameters of left-right variance And The specific formula is as follows: (11) the peak value after compensation is deduced through the judgment of variance: (12) Wherein F' is the peak value preliminarily determined in step S3, and F is the compensated peak value.
2. The asymmetric gaussian model-based fiber bragg grating signal peak detection algorithm according to claim 1, wherein the five-point sliding mean filtering method in step S1 has a calculation formula: (1) Where n is the number of data points, Xi represents the abscissa of the ith point and yi represents the ordinate of the ith point.
3. The algorithm of claim 1, wherein the size of the window in step S2 is the difference between the abscissa where the minimum point and the maximum point of the derivative are located.

Description

Optical fiber Bragg grating signal peak detection algorithm based on asymmetric Gaussian model Technical Field The invention belongs to the field of fiber gratings, and particularly relates to an optical fiber Bragg grating signal peak detection algorithm based on an asymmetric Gaussian model. Background The fiber Bragg grating (Fiber Bragg Grating, FBG) sensor is widely applied to health and security monitoring of mines, bridges, dams and composite structures due to the advantages of small volume, good stability, high precision, strong electromagnetic interference resistance, corrosion resistance, low cost, passive intrinsic safety and the like. The fiber Bragg grating sensor reflects the measured change by acquiring the center wavelength drift, and the reflection spectrum peak value position corresponding to the wavelength drift changes, so that the demodulation accuracy of the wavelength of the fiber Bragg grating is significant. The fiber grating sensor is widely applied to the engineering and industrial fields, and the traditional fiber grating demodulation algorithm has the defects of low precision, low running speed and poor noise resistance, and can not meet the requirements of a high-precision and real-time dynamic demodulation system. Factors such as light source, multiplexing technology, noise, nonlinear temperature drift of a measuring device, spectrum distortion and spectrum overlapping caused by external environment are important reasons for low demodulation precision of the fiber bragg grating sensor, and improvement of the problems becomes a research hot spot in recent years. At present, common peak searching algorithms include a direct peak searching (DP) method, a polynomial fitting method, a Gaussian fitting method, a three-point peak searching method, a genetic algorithm and a neural network algorithm. The DP method is easy to operate, but has higher requirements on the sampling points of the spectrum, the portable demodulator generally has fewer sampling points, the demodulation precision of the DP method is lower, the operation amount of the polynomial fitting method is small and easy to realize, but the peak detection data precision of the method mainly depends on observed data, the Gaussian fitting algorithm finds the peak point through the symmetric relation of the left side and the right side of a reflection spectrum signal, the requirements on the spectrum shape are strict, and when the reflection spectrum shape is distorted due to noise, the peak detection error of the algorithm is increased. The three-point peak searching method has a certain improvement on peak searching accuracy, but does not consider the asymmetric characteristic of spectrum peaks in the peak searching process, and the existing FBG spectrum peak searching algorithm is subjected to analysis and research from the aspects of detection accuracy, noise resistance and the like of the algorithm, so that the influence on the spectrum asymmetry is less. The FBG reflection spectrum is a nonstandard Gaussian spectrum, and the wave crest shape is irregular. Therefore, the problem of FBG asymmetric spectrum peak searching is to be further researched, and the method has important significance for perfecting a peak searching algorithm and improving detection accuracy. Disclosure of Invention In view of this, the present invention proposes an optical fiber bragg grating signal peak detection algorithm based on an asymmetric gaussian model, comprising the steps of: s1, smoothing the collected optical fiber Bragg grating spectrum signal by adopting a five-point sliding average filtering method to obtain a smoothed spectrum signal; s2, conducting primary derivation on the spectrum signal after the smoothing processing, and determining the size of a window according to the positions of the maximum value point and the minimum value point of the derivative; S3, resampling the spectrum signals in the window, reducing the sampling rate, adopting Gaussian fitting to the resampled spectrum signals in the window, and preliminarily determining the position of the peak value; S4, introducing an asymmetric Gaussian function model, and correcting the initially determined peak position. Further, in step S1, the calculation formula of the five-point sliding average filtering method is as follows: Where n is the number of data points, i=1, 2Λ, n, x i represents the abscissa of the i-th point, and y i represents the ordinate of the i-th point. Further, the size of the window in step S2 is the difference between the abscissa where the minimum point and the maximum point of the derivative are located. Further, the step S3 specifically includes: S31, the fiber Bragg grating signal is expressed as: Wherein lambda is the wavelength of the spectrum of the fiber Bragg grating, lambda B is the center wavelength of the spectrum of the fiber Bragg grating, delta lambda B is the 3dB bandwidth, and A is the amplitude of the reflection spectrum; Taking the