The invention discloses a Gaussian multi-peak fitting-based spectrum analysis algorithm. The Gaussian multi-peak fitting-based spectrum analysis algorithm comprises the following steps : determining an absorption fitting function containing an undetermined Gaussian feature parameter, performing mathematical modeling on fitting error, determining center coordinate and iteration step of an iteration initial value, continuously performing iteration, performing comparison and changing the iteration step length until the optimal fitting error is obtained, and reversely deducing the center coordinate value, namely the to-be-determined feature parameter, of the corresponding division unit by the optimal fitting error, and determining the absorption fitting function so as to decompose the mixed spectrum into a plurality of sub-peaks. The Gaussian multi-peak fitting-based spectrum analysis algorithm has the following advantages : a Gaussian function curve and a spectroscopic data curve have high affinity and can express parameters, such as peak shape, peak height and peak position in the spectrogram, with clear physical significance; and the mixed absorption spectrum curve is decomposed into a plurality of Gaussian function curves, the effective information of the original spectrum data can be remained, the original spectrum data is unified into a small amount of uniquely determined Gaussian feature parameters, and the aims of simplifying the spectrum data and researching the refined structure are fulfilled.
; 王焕卿
修改评论