Research On Decomposition Method Of Mixed Spectrum Of Soil Clay Mineral

With the rapid development of economy and accelerate industrialization process, the demand for mineral resources increases day by day,short of mineral resources not only restrict the sustained and stable economic development, but also relate to the security of national resources. To improve the efficiency of exploration, remote sensing information extraction technology for rock is an important approach which has been widely applied to resource exploration.Mixed spectral decomposition method is an important part of remote sensing information extraction technology for rock and mineral. Mixed spectral decomposition method is a calculating method based on the spectra-physics- mathematical model. The establishment of spectra-physics-mathematical model began from the mechanism of spectrum, when the light goes into the interior material, will produce the behavior of absorption, this absorption of such behavior is the spectral response of internal structure of matter, trace elements and characterization types. Just as material has this physical mechanism, so starting from the physics of spectrum, the quantitative model of the extraction of material information is possible.Just as a kind of electromagnetic wave, light has wave-particle duality. According to electromagnetic theory of matter, the production of all matter spectrum has strict physical mechanism, as clay minerals in soil has its inherent characteristics of reflection, absorption, transmission and radiation of electromagnetic wave, so it has its own unique spectral characteristics. Spectral characteristics of clay minerals in soils is mainly composed of components, material within the crystal structure, the influence of physical and chemical characteristics, generally, spectral characteristics of clay minerals in soils is relatively stable. Therefore, the spectral characteristics of clay minerals in soils can be differentiated.Soil clay minerals in the near infrared spectrum range has diagnosis. Near infrared spectroscopy is vibrational spectra of molecular spectra, it is the fundamental frequency of molecular vibration frequency multiplier and cooperation, including a large number of functional groups characteristic information. Modern near-infrared spectrometer is fast on the test samples, and is the measurement of preventing samples and testers. Near-infrared spectral data collected up has nm sample interval, but rich data also brings redundancy, resulting in the information efficiency of near-infrared spectral data is low. So enhanced the characteristic information, mainly used the envelope removed method to normalize the background value, enhanced the information of near-infrared spectral data, searched the absorption peak position and quantified the information, including depth, width of absorption, absorption symmetry and the extraction of the absorption area, and in the end, got near infrared spectral data on feature space optimization.The mixed spectrum model of clay minerals in soil includes linear model and nonlinear model, using programming software to realize matrix operations, with binding of the least squares and linear programming for solution based linear model, and the nonlinear constrained optimization based nonlinear model, inverse the content of soil clay minerals, and have the verification with real content of the known mineral phase, screening algorithm.This research provides a feasible method for near-infrared spectral information characteristic extraction and retrieval of soil clay minerals, and offers a new direction and extension for the practical application of spectroscopy. Near-infrared spectroscopy as a hyper spectral remote sensing rock extraction technology, is able to develop its skill to full in terms of soil clay quantitative inversion.