Rapid Monitoring of Key Indicators in Growth Process of Chlorella Using Near-Infrared Spectroscopy Technology
10.19756/j.issn.0253-3820.241492
- VernacularTitle:基于近红外光谱技术快速监测小球藻生长过程中的关键指标
- Author:
Wen-Hui SONG
1
;
Shi-Jie DU
;
Yan LIU
;
Qiao WANG
;
Xin LIU
;
Zhi-Yong GONG
Author Information
1. 武汉轻工大学食品科学与工程学院,武汉 430023
- Keywords:
Near infrared spectroscopy;
Chlorella;
Spectral pretreatment;
Wavelength screening;
Partial least squares regression
- From:
Chinese Journal of Analytical Chemistry
2025;53(4):660-668
- CountryChina
- Language:Chinese
-
Abstract:
The traditional detection methods for monitoring the biomass,protein,chlorophyll content and other key indicators in the growth of chlorella have some problems,including complicated operation,slow detection speed and difficult large-scale application.In this study,a fast and efficient monitoring method for the key indicators in the growth of chlorella was established using near infrared spectroscopy and chemometrics.Near-infrared spectroscopy was used to collect near-infrared spectra of chlorella algal fluid at different growth stages,and standard methods were used to detect the biomass,protein and chlorophyll contents of corresponding samples.A quantitative analysis model was established based on partial least squares regression(PLSR).To improve the prediction ability of the model,multiplicative scatter correction(MSC)was used to reduce the interference of scattering on the raw spectrum(RS),standard normal variate(SNV)was used to normalize the original spectral data to eliminate differences between samples,continuous wavelet transform(CWT)was used to obtain the key features of spectral data,the first derivative(1st)was used to enhance the differentiation of the original spectral features,and monte carlo-uninformative variable elimination(MC-UVE)and randomization test(RT)were used to screen the valid variables in the wavelength.By evaluating the prediction ability of different models,the quantitative analysis models of chlorella biomass,protein and chlorophyll content were finally determined.The results showed that the model based on 1st combined with RT spectra had better predictive ability for chlorella nutrient content detection,and the root mean square errors of prediction(RMSEP)and coefficients of determination(R2)were 0.041 and 0.933 for biomass,0.012 and 0.973 for protein,and 0.517 and 0.962 for chlorophyll,respectively.This model showed practical application value,and could realize the rapid and accurate detection of chlorella biomass,protein and chlorophyll content at the same time.
