Research Of Several Key Technologies For Environment Monitoring System Based On FTIR

Environmental monitoring has become the focus of attention. Thedevelopment of environmental monitoring technology exerts an importantsignificance to environmental pollution monitoring, control and governance. Due tohigh resolution, high signal-to-noise ratio, high-throughput, high measurementaccuracy as well as research spectral range, Fourier Transform InfraredSpectroscopy (FTIR) is widely used in indoor and outdoor atmosphericenvironmental monitoring.The paper focuses on the key technologies of the environmental monitoring.And the main research work and achievements of this dissertation are listed asfollowing:(1) We analyzed the current situation and existing problems of FTIRtechnology at home and abroad andidentified the principles of the use ofapodization function. We conduct research for2N FFT algorithm and write theinterferograms data processing software.(2) To tackle phase error exist in the interferograms, we improve the all-passfilter phase correction method by design the best low-pass filter toapproximatehigher-order all-pass filter of a given phase response theinterferograms phase error.it is verified by experiment, the improved method obtaina better correction effect.(3) To tackle the problem of the noise and baseline drift which exist in themeasured spectrum.The several common wavelet transformation methods have beenadopted to achieve the pretreatment of spectral signal. Experimental verificationadvantages and disadvantages of various de-noising method, the signal-to-noiseratio (SNR), root mean square error (RMSE), waveform similarity (NCC),smoothness (r) four indicators are used to evaluate the system experimentalverification of the wavelet de-noising and baseline correction.(4) To make up for the shortcomings of the wavelet transform in losing usefulsignal, a new method based on adaptive sparse representation has been proposed todeal with the measured spectral signal, the results it can solve the problem of the loss of useful information in the process of wavelet de-noising.