Improvement of NIR models for quality parameters of leech and earthworm medicines using outlier multiple diagnoses

Leeches and earthworms are the main ingredients of Shuxuetong injection compositions,whichare natural biomedicines.Near infrared(NIR)diffuse reflection spectroscopy has been used forquality assurance of Chinese medicines.In the present work,NIR spectroscopy was proposed as arapid and nondestructive technique to assess the moisture content(MC),soluble solid content(SSC)and hypoxanthine content(HXC)of leeches and earthworms.This study goal was toimprove NIR models for accurate quality control of leech and earthworm using outlier multiplediagnoses(OMD).OMD was composed of four outlier detection methods:spectrum outlier di-agnostic(MD),leverage diagnostic(LD),principal component scores diagnostic(PCSD)andfactor loading diagnostic(FLD),Conventional outlier diagnoses(MD,LD)and OMD werecompared,and the best NIR models were those based on OMD.The correlation coefficients(R)for leech were 0.9779,0.9616 and 0.9406 for MC,SSC and HXC,respectively.The values ofrelative standard error of prediction(RSEP)for leech were 2.3%,5.1%and 9.0%for MC,SSC and HXC,respectively.The values of R for earthworm were 0.9478,0.9991 and 0.9605 for MC,SSC and HXC,respectively.The values of RSEP for earthworm were 8.8%,2.4%and 12%for MC,SSC and HXC,respectively.The performance of the NIR models was certainly improved by OMD.

A novel multiple-outlier-robust Kalman filter

This paper presents a novel multiple-outlier-robust Kalman filter(MORKF)for linear stochastic discretetime systems.A new multiple statistical similarity measure is first proposed to evaluate the similarity between two random vectors from dimension to dimension.Then,the proposed MORKF is derived via maximizing a multiple statistical similarity measure based cost function.The MORKF guarantees the convergence of iterations in mild conditions,and the boundedness of the approximation errors is analyzed theoretically.The selection strategy for the similarity function and comparisons with existing robust methods are presented.Simulation results show the advantages of the proposed filter.