论文部分内容阅读
目的本研究基于蚁群-遗传区间偏最小二乘(ACO-GA-iPLS)近红外谱区筛选方法预测花茶花青素含量。方法首先对花茶近红外光谱进行预处理;然后用ACO-iPLS优选出特征子区间;最后对所选的特征子区间,用GA-iPLS进一步细化花青素的特征子区间,并建立花青素的预测模型。结果优选出3个特征子区间(第1、9、10子区间),所建模型对应的交互验证均方根误差(RMSECV)和预测均方根误差(RMSEP)分别为0.1460mg/g和0.1840 mg/g,校正集和预测集相关系数分别为0.9187和0.8856。结论ACO-GA-iPLS可以有效选择近红外光谱特征波长,简化模型,提高模型精度。
OBJECTIVE In this study, anthocyanin-genetic interval partial least squares (ACO-GA-iPLS) near infrared spectral screening method was used to predict the content of anthocyanin in Camellia. The method was applied to pretreatment of near-infrared spectra of tea plants. Then, the characteristic subsets were optimized by ACO-iPLS. Finally, the characteristic subsections of anthocyanins were further refined by GA-iPLS for the selected characteristic subsections, Predictive model of prime. Results The results showed that the RMSECV and RMSEP of the proposed model were 0.1460mg / g and 0.1840 respectively mg / g, the correlation coefficient between the calibration set and the prediction set are 0.9187 and 0.8856, respectively. Conclusion ACO-GA-iPLS can effectively select the characteristic wavelength of near-infrared spectroscopy, simplify the model and improve the precision of the model.