Abstract
During the online quantitative analysis of total nitrogen, numerous data is needed for dictionary learning to better represent the spectrum. However, the amount of on-site sampling data is limited, which cannot meet the requirement of dictionary learning. Therefore, this paper proposes a pre-processing algorithm based on the transfer dictionary, which can update the source dictionary with small samples to perform online spectral pre-processing. To better measure the effectiveness of the method, the experiment executes classical pre-processing algorithms separately, and then combines with partial least squares to predict the concentration of total nitrogen. The results show that the method proposed in this paper can improve the accuracy of total nitrogen detection.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call