An analytical method in a quasi-static fiber grating sensing system under transverse uniform press is proposed based on genetic algorithm The effect of population size, generations, crossover ratio and mutation ratio to genetic algorithm, and the optimization parameters of genetic algorithm were given. The relevant experimental system is constructed. The simulation and experiments show that the analytical method proposed can be applied to analyze the reflective spectra of the quasistatic FBG sensing system at transverse uniform press, the strain measurement with high-precision of 0.91% can be realized CLC numbers: TN253 Document code: A

The integration of renewable energy generation and nonlinear power electronic equipment into the grid brings about complex harmonics and interharmonics problems. The amplitude and frequency of harmonics and interharmonics should be detected by high time-frequency (T-F) resolution methods owing to their time-varying transient features. In this paper, a synchrosqueezing adaptive S-transform (SAST) method is proposed to detect the parameters of harmonics. Firstly, the time-frequency spectrum (TFS) of the harmonic signals is acquired by an adaptive S-transform (AST) algorithm. The TFS results are then subjected to synchronous compression, so as to achieve higher time-frequency representation precision. The detection results of the simulation signals show that SAST can achieve a better time-frequency resolution than the S-transform (ST) and synchrosqueezing short-time Fourier transform (SSTFT). In addition, the application of SAST to the analysis of experimental signals also suggests its superiority in the parameter detection of harmonics, especially for the time-varying interharmonics.

Fundamental optical constants of porous silicon, which refer to the effective refractive index n eff , the extinction coefficient K, the real parts ε er and imaginary parts ε ei of the effective dielectric functions, and the absorption coefficient α, are calculated using the effective medium theory named the Modified two-phase Maxell-Garnett and two-phase Bruggeman models separately in the range of 400-1000 nm. In the model, the microstructure of porous silicon is considered as a two-phase granulated compound medium with identical silicon inclusions, placed at random inside a homogenous air matrix. To do the calculation the calculated reflectance spectra from two models are fitted with experiments corresponding to samples with porosities of 39%, 45%, 51%, 59%, and 67% in order to determine the involved parameters in dielectric function called Wemple-DiDomenico oscillator of the remaining silicon. Best results of both models are obtained using spheroidal inclusions of 0.9 eccentricity. Effects of the wavelength of incident light and the porosity of porous silicon on these optical constants were studied systematically.

A modified genetic algorithm (GA) has been proposed, which was used to wavelength demodulation in quasi-static fiber grating sensing system. The modification method of GA has been introduced and the relevant mathematical model has been established. The objective function and individual fitness evaluation strategy interrelated with GA are also established. The influence of population size, chromosome size, generations, crossover probability and mutation probability on the GA has been analyzed, and the optimal parameters of modified GA have been obtained. The simulations and experiments, show that the modified GA can be applied to quasi-static fiber grating sensing system, and the wavelength demodulation precision is equal to or less than 3 pm. CLC numbers: TN253 Document code: A Article ID:Genetic algorithm (GA) was an optimization algorithm based on imitation to biological evolution mechanism. As an adaptive heuristic global searching algorithm, it has been used widely in many fields [1,3] . However, the FBG demodulation system used widely at present was fiber F-P demodulation system whose demodulation precision was only 5 pm [4] . In order to improve the demodulation precision, a method that combines an modified GA with fiber grating demodulation technique was proposed. Moreover, the modified GA was applied in quasi-static fiber grating sensing system and the corresponding experiments have been done. It has been shown from the experiments that the system demodulation precision is less than 3 pm.The typical fiber grating sensing system has been shown in Fig.1, in which the reflective signals were demodulated by the tunable F-P filter.*

A novel fiber grating vibration demodulation system, based on 2×2 and 3×3 couplers, is designed. Based on the phase unwrapping algorithm, the three-way asymmetrical output of the 3×3 coupler and demodulation state characteristics of the system when the vibration signal includes high harmonic waves are analyzed in simulations. The result shows that when the three-way output is asymmetrical, the maximum deviation of demodulation signal is 1.625%, and when the vibration signal includes high harmonic waves, the maximum deviation of demodulation signal is 0.9%. The corresponding experiment is conducted. The experimental result shows that the dynamic resolution of the system is 25.22Hz / n when the vibration pitch is 5.5 Hz.