近红外特征波长筛选在勾兑梨汁中原汁含量的快速检测中的应用

为实现近红外光谱进行勾兑梨汁中原汁含量的快速检测，采用相同可溶性固形物含量的新鲜梨汁和果汁粉冲剂按照原汁含量为0%～100%进行勾兑，并结合遗传算法(GA)、粒子群算法(PSO)以及萤火虫算法(GSO & FA)进行特征波长筛选，比较分析四种算法分别建立的偏最小二乘(PLS)模型。结果表明，GA-PLS，PSO-PLS，GSO-PLS，FA-PLS四种模型均能够剔除大部分波长变量，其中以FA-PLS模型效果最佳，不仅保证模型的稳健性，而且简化了模型，提高了预测的精度。为了进一步优选特征波长，利用连续投影算法(SPA)在FA基础上做进一步波长筛选，并比较全波段PLS，SPA-PLS，FA-PLS，FA-SPA-PLS模型，四种模型泛化能力为：FA-PLS>PLS>FA-SPA-PLS>SPA-PLS，其预测均方根误差分别为0.029 1，0.033 3，0.033 9和0.137 0，相应的波长变量数量依次367，765，20和18。其中SPA-PLS波长变量最少，但预测误差远远高于其他三种模型，综合考虑预测精度与波长变量数目，FA-SPA-PLS模型不仅波长变量较少而且预测精度较高，能够有效鉴别勾兑梨汁中原汁含量。研究利用近红外光谱技术为快速鉴别勾兑果汁提供一种有益思路，并通过波长变量筛选简化定量分析模型。

Application of Characteristic Wavelength Variable Application of NIR Spectroscopy Based on Swarm Intelligence Optimization Algorithms and SPA in Fast Detecting of Blending Pear Juice

In order to rapidly determine the content of raw juice in blending pear juice by near-infrared spectroscopy (NIR),ex-periments using the same soluble solids content of fresh pear juice and juice powder were conducted.Four common swarm intelli-gence optimization algorithms,including Genetic Algorithm (GA),Particle Swarm Optimization (PSO),Glowworm Swarm Op-timization (GSO)and Firefly Algorithm (FA),were combined with PLS to select wavelength variables.The results showed that the four kinds of models could remove most of the wavelength variables,and the FA-PLS model achieved the optimal perform-ance,which simplified the model and improved the accuracy of prediction.Then,the successive projections algorithm (SPA) was used to select wavelength variables after Firefly Algorithm (FA).The results indicated the generalization ability were as fol-low:FA-PLS>PLS> FA-SPA-PLS>SPA-PLS.The root mean square errors of prediction (RMSEP)was 0.0291,0.0333, 0.0339,0.1370,respectively,and the corresponding wavelength variables number were 367,765,20,18.The wavelength variables of SPA-PLS model were the least,but RMSEP was much higher than the other three models.Considering the predic-tion precision and the number of wavelength variables,the FA-SPA-PLS model was validly improved with less wavelength varia-bles and higher prediction accuracy.This study provides a convenient way for rapid identification of blending fruit juice using NIR.