Research On Surface Plasmon Resonance Sensor Based On Dielectric-aided Layer And Its Spectrum Signal Processing Method

Surface plasmon resonance (SPR) effect is a physical optics phenomenon which takes place onthe interface of metal-dielectric medium, involving charge oscillations excitated by attenuated totalreflecting. For the capacity of resisting disturbance, no labeling and low response time, it has beenwidely used in many fields, such as biochemical medicine, environmental monitoring. In order tohence the sensitivity of the SPR sensor and extend the application of SPR technology, the profoundresearches are implemented theorietically and experimentally on the incentive model, the spectrumsignal processing. This paper research content is divided into the following several parts:Firstly, the principle and conditions of SPR effect basd on the Kretschmann model is researched.Numerical simulation of internal and external dielectric-aided layer prism SPR incentive model wascarried out by using the finite difference time domain method for resonance excitation characteristics.Theoretical basis for improving the detection sensitivity and extend its application field are providedby exploring the SPR resonance excitation performance influence factor.Secondly, the prism SPR sensor model based on prism-dielectric-aided layer-gold-environmentalmedia is presented, and the wavelength modulation SPR excitation system is built up. Dielectric-aidedlayer can effectively improve the sensitivity of detection prism SPR sensor, and realizes the effectivemodulation of the resonance spectra.Thirdly, a metal corrosion detection based on optical fiber SPR sensor is proposed. By studyingthe relationship of SPR sensor reflection spectrum and the corrosion state, the corrosion rate anddegree can be established so that the SPR technology can be provide a basis for the application in thefield of corrosion monitoring.Finally, the method of the wavelet packet transform based on EMD is proposed. The noisesimulation resonance signal and actual measurement income resonance spectrum signal are studied.The result shows that the method can not only filtered out the noise from the signal effectively, butalso can extracted resonance spectrum feature information accurately. Identification precision of theresonance trough can be significantly improved.