Research On Vibration Sensing And GPU Accelerated Processing Based On OFDR

Distributed optical fiber sensing technology has become an indispensable part in the field of optical fiber technology.Compared with traditional electromagnetic sensors,it has the advantages of strong anti-electromagnetic interference ability,high electrical insulation,high chemical stability,and high sensitivity,and has attracted more and more attention and research.Due to its high sensitivity and high spatial resolution,optical frequency domain reflection technology has broad prospects and markets in aerospace,medical and health,engineering testing and precision instrument manufacturing industries.This paper mainly aims to build OFDR real-time sensing system,and conducts research on OFDR vibration sensing algorithm and GPU acceleration processing rate.The basic content of this article is as follows:1.Based on a detailed study of the basic principles of OFDR technology,Rayleigh scattering,coherent detection and other theories are studied in detail.The influence of various parameters in OFDR system is studied,and the basic requirements of OFDR system components are put forward.The processing algorithm of OFDR sensing system is further explained.2.The nonlinear effect of the tunable light source is analyzed?The hardware and software compensation methods are proposed respectively,focusing on the cubic spline interpolation compensation algorithm in software compensation and the formula derivation is carried out.The effectiveness of the hardware and software compensation methods are verified by experiments respectively,and the advantages and limitations of the hardware and software compensation methods are compared.In this experimental platform,the hardware method is selected as the compensation method for nonlinear effects.3.The OFDR vibration sensor algorithm is introduced.An OFDR vibration sensor algorithm based on sliding time gating is proposed.The differences between the two algorithms are compared,and the effect of the new algorithm on the frequency response range of the OFDR vibration sensor system is analyzed mathematically.4.Experimental research on OFDR vibration sensing is carried out.Selecting suitable lasers and detectors,an OFDR sensing system based on hardware compensation is built,and the experimental structure is optimized.To add vibration signals by PZT,two algorithms are used to process them respectively to verify the accuracy of the moving time-gating algorithm and realize the phase analysis of 50Hz and 100Hz vibration signals after the vibration signals are collected.Under the same spatial resolution,the system frequency response range is increased from 50Hz to 100Hz.At the same time,the acquisition of multi-frequency signals is realized.5.The basic background of GPU acceleration is introduced and the parallel characteristics of OFDR processing algorithms are analyzed.The fast Fourier transform(FFT)and cross-correlation operations of the key parts of the OFDR processing algorithm are analyzed.The algorithm parallel processing is performed,and GPU is used to accelerate the calculation.The overall OFDR algorithm is analyzed and parallelized.The conversion of the OFDR processing platform from CPU to GPU is realized,and the processing speed is increased from 1017ms to 45.5ms,which is about 20 times faster.The innovation of this paper:1.The OFDR vibration sensing algorithm is optimized.Based on the original OFDR vibration sensing algorithm,the concept of moving time-gating is introduced,which significantly improves the frequency response range of vibration sensing.2.GPU is used to accelerate the calculation of OFDR static sensor,and the OFDR static processing algorithm is designed in parallel to improve the processing rate of OFDR system.