Research On Self-mixing Interferometry Singal Noise Processing

When a beam of light emitted from a laser source is reflected or scattered back into the laser cavity by an external target,the reflected light will mix with the light inside the cavity,causing a modulation of the laser output power.So it is called optical feedback self-mixing interference(OFSMI).The laser self-mixing interference system has a very simple and compact set-up and auto collimation capability.Because of carrying external reflector information,the system can be used for multi physical quantity sensor testing,especially suitable for sensing measurement of mechanical quantities such as displacement,vibration,distance,contour.Light interference measurement with high resolution,high precision characteristics,which makes the optical feedback self-mixing interference technology has broad prospects in the sensing measurement field.In order to verify the correctness of the theoretical model,the self-mixing interference signals in different feedback levels are simulated by Matlab basing on the theoretical model of self-mixing interference.Therefore,the study focuses on the research of two kinds of method to eliminate noise in the OFSMI signal.One is based on wavelet transform processing.Wavelet transform decomposed different kinds of frequency element to non-overlapped frequency bands,and this method puts forwards an effective way to signal filter,signal-noise separating and character picking-up,especially it has well de-noising capability.Through the simulation experiment,the signal-to-noise ratio is improved greatly.The other one is based on singular value decomposition processing.The phasespace reconstruction of the measurement signals of OFSMI was established,and then the attractor matrix was decomposed by using SVD for de-noising.Through the simulation experiment,the SVD method can effectively reduce the noise component in the signals.Finally,conclusions are drawn from the whole study,and the further research issues are also pointed out.