Font Size: a A A

Research And Implementation Of The Demodulation Algorithm For Fiber Fabry-Perot Sensors

Posted on:2022-02-13Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y WuFull Text:PDF
GTID:1488306572474994Subject:Optical Engineering
Abstract/Summary:PDF Full Text Request
The Fabry-Perot Interferometry(FPI)sensor has the characteristics of small size,high measurement accuracy,good stability,which is widely used in the field of static parameter sensing,such as measuring the strain of the wing to monitor the fatigue of the aircraft status,health monitoring of large structures such as bridges and dams.The FPI sensor can also be used in the chemical industry to monitor the refractive index and temperature of reactants in real time.The demodulation accuracy and sensitivity determine the detection accuracy of the FPI sensor.The coverage of the sensing system is limited by the multiplexing capacity of the sensor.The sensor may also suffer from the influence of temperature fluctuation in practical applications.This paper analyzes the existing demodulation algorithm of fiber FPI sensor in detail,and points out the shortcomings of the existing demodulation algorithm.It focuses on the demodulation accuracy,sensitivity,multiplexing method and antitemperature interference.The main research contents and results of the thesis are summarized as:? A demodulation algorithm for Fabry-Perot sensors using lasers for coarse spectral sampling and reconstruction is firstly proposed and demonstrated.A tunable laser is used to coarsely sample the spectrum of the Fabry-Perot sensor.Maximum likelihood method is used to estimate the sensor's parameters according to the theoretical spectrum.The cavity length of the sensor can be demodulated through fitting the sampled data and the theoretical spectrum.Theoretical analysis and simulations indicate that the minimum cavity length resolution of the demodulation algorithm depends on the sampling interval and the signalto-noise ratio.Experimental results show that the demodulation algorithm has a large cavity length demodulation range.In the repeated strain measurement,the cavity length demodulation fluctuation range is less than 0.84 nm and the standard deviation is 0.159 nm.? A highly sensitive FPI sensor demodulation method based on spectral sampling Vernier effect is proposed and demonstrated.In order to produce the vernier effect,the sampling interval of the tunable laser is very close to the free spectral range of the FPI sensor.The spectrum obtained by coarse sampling is similar to the spectrum envelope in the Vernier effect.The spectral shift can be demodulated by tracing appropriate envelope peak wavelength.In the strain sensing experiment,the sensitivity is amplified by 14 times compared with the single peak finding method,reaching 0.01836 nm/??.By changing the sampling interval,the demodulation sensitivity and the corresponding measurement time can be flexibly adjusted.For the existing FPI sensors network,the online upgrade of the sensor system can be realized by replacing the system light source with a tunable laser.? A method for the FPI sensors multiplexing based on the neural network is proposed.The FPI sensors with different cavity lengths have different spectral responses to strain.The neural network is used to find the relationship between the sensor spectrum and the strains applied to each sensor.The well-trained network is used to demodulate the spectrum and extract strain information.In the experiment,sensors with cavity lengths of approximately 160 ?m and 180 ?m were fabricated.The spectra of the two sensors under different strains were collected and synthesized as the training data.The well-trained network is used for demodulation.The crosstalk of the feedforward neural network is as low as 4.31 ?? in the range of 0-1020 ??,while the Fourier transform method shows serious crosstalk because of the sidelobes.Theoretical analysis shows that the demodulation performance of the neural network mainly depends on the signal-to-noise ratio of the measuring spectrum rather than the training spectrum.? An temperature insensitive refractive index sensor is proposed and demonstrated.A hybrid FPI structure consists of an air cavity and an intrinsic cavity is designed and fabricated.The power responses of two lasers are measured by a dual-channel optical power meter simultaneously.The two reflected power signals distribute along an ellipse.The refractive index of the liquid is calculated from the semi-major axis of the fitted ellipse.The refractive index sensing system is used to measure the refractive index of the salt solutions with different concentrations.The demodulated results matched well with the refractive index measured by the Abbe refractometer,and a resolution of 0.0017 was obtained.Since the temperature is eliminated during the ellipse fitting,the measuring result is insensitive to the temperature fluctuation.The refractive index of pure water from boiling cooling to room temperature is measured by this sensor,which proves that the proposed sensor is insensitive to temperature fluctuation.
Keywords/Search Tags:Fiber optic sensor, Phase modulation, Fabry-Perot interferometer, Signal demodulation, Wavelength sampling
PDF Full Text Request
Related items