Research Of The Key Techniques Of DMD Based Hadamard Transform Near-infrared Spectrometer

Component analysis is an important means which used by scientists and the public to realize the world. As an important technology of ccomponent analysis, spectral analysis is widely used in clinical medicine, scientific research, food testing, petrochemical production, space remote sensing, environmental measurement and other fields because of the advantages such as rapid, non-destructive and real-time detection. As an important part of the world, the well-known organic compound is up to 9 million. The spectral analysis of organic is concentrated in near-infrared range. Because of the weakness of near-infrared spectra, the signal noise ratio of NIR spectrometer declines. Then the Hadamard transform(HT) algorithm is introduced to the near-infrared(NIR) spectrum detection. HT near-infrared spectrometer is a new kind of digital transform spectrometer, which based on combination detection of spectrum detection. Compared to conventional spectrometers, Hadamard transform spectrometer increases the number of entrance(exit) slists from one to multiple so that the energy detected by detector increases and the SNR is improved. As an ideal spectrum strobe component of HT spectrometer, Digital Micro-mirror Device(DMD) has several advantages including more flexiblity and stabilty of DMD. DMD-HT spectrometer has been studied extensively in recent years. Combined the application in organic compound component analysis, the DMD-HT near-infrared spectrometer is analyzed and studied in our group.Firstly, the spectral culvature is analyzed and modified in DMD-HT spectrometers. In DMD-HT spectrometer, the monochromatic images of slit are curved and the spectrum is aliasing because of the diffraction of plane grating, the design defect of optical system and the consideration of aberration balance. Eventually, the resolution of the spectrometer declines. Firstly, we make the encode stripe coincidence with the spectral stripe by adjusting the DMD. Then we correct the remained spectrum aliasing by introducing a correction matrix. Simulation results show that the spectral resolution is improved gradually by adjusting the encode stripes and processing data.Secondly, the concentrator system in DMD-HT spectrometer is redesigned. In DMD-HT spectrometer, the spectrum is of crosstalk because of the diffraction of DMD and stray light in spectrometer. Then the SNR in DMD-HT spectrometer declines. Firstly, the diffraction of DMD and stray light of spectrometer are analyzed. Then, the concentrator system is redesigned by using CPC(compound parabolic concentrator) based on non-imaging optic design. Simulation results show that the concentrator system with CPC has a good performance both on spectrum concentration and stray light suppression and the CPC-based concentrator system has a large tolerance. Similarity,the spectrum concentration efficiency in DMD-HT spectrometer with double sub-gratings(DSG) is improved highly when the concentrator system is design with CPC.Thirdly, the spectrogram distortion is corrected in DMD-HT spectrometer. In DMD-HT spectrometer, the spectrogram is distorted because of the stray light and spectral response deviation in spectrometer. The factors are analyzed separately. On the one hand, the stray light in DMD-HT spectrometer is analyzed and classified. Firstly, the stray light is devided into variable stray light and intrinsic stray light Then, the impact of stray light on spectrum detection is analyzed in the view of encode equation. Finally, the decode equation eliminating stray light is derived and the spcectrum is corrected by processing data. The simulation results show that the performance of stray light elimination algorithm in spectrum correction is minimal but the accurate coverage of spectral absorbance is improved highly. On the other hand, the spectral response divation in DMD-HT spectrometer is analyzed and classified. Firstly, the spectral response divation is devided into time-related spectral response divation and wavelength-related spectral response divation. Then, the impact of spectral response divation on spectrum detection is analyzed in the view of encode equation and the decode equation eliminating spectral response deviation is derived. Finally, the spectral response deviation is ascertained by inserting test template during the encoding process and testing wavelength-related spectral response divation before encoding process. Experimental results show that the spectral response deviation correction method is simple and effective. And the spectral response deviation correction method is suitable in DMD-HT spectrometer with double sub-gratings simillarly.Finally, the spectrum acquisition electrical system and spectrometer specific software of DMD-HT spectrometer are designed and the stability test of prototype is carried out. In DMD-HT spectrometers, the HT encoding process is carried out by controlling the DMD and the decoding process is carried out by specific spectrometer software. The suitable spectrum acquisition process of DMD-HT spectrometer has great influence on the performance of spectrometer. After analyzing the spectrum acquisition process, the electrical structure of DMD-HT spectrometers is redesign and the spectrometer software is written. Finally, the stability of prototype is tested for DMD-HT near-infrared spectrometer entering the market.