种荧光光谱法和PSO-BP神经网络相结合的山梨酸钾浓度测定的新方法（英文）  

A New Method for the Determination of Potassium Sorbate Combining Fluorescence Spectra Method with PSO-BP Neural Network

作　　者：王书涛[1] 陈东营 王兴龙[1] 魏蒙[1] 王志芳[1] 

机构地区：[1]燕山大学电气工程学院,河北省测试计量技术及仪器重点实验室,河北秦皇岛066004

出　　处：《光谱学与光谱分析》2015年第12期3549-3554,共6页Spectroscopy and Spectral Analysis

基　　金：National Natural Science Foundation of China(61201110)

摘　　要：研究了山梨酸钾在水溶液和橙汁中的荧光特性,结果表明在两种溶液中山梨酸钾的荧光特性虽然有很大的区别,但是它们的荧光特征峰都存在于λex/λem=375/490nm。从二维荧光光谱可以看出,橙汁中山梨酸钾的浓度和相对荧光强度关系错综复杂,两者不再满足线性关系。为了准确测定橙汁中山梨酸钾的浓度,提出了一种微粒群（PSO）算法优化的误差逆向传播（BP）神经网络的新方法。两组预测浓度的相对误差分别为1.83%和1.53%,预测结果表明该方法具有可行性。在浓度范围为0.1~2.0g·L-1内,PSO-BP神经网络能够完成橙汁中梨酸钾浓度的准确测定。In this paper, fluorescence spectra properties of potassium sorbate in aqueous solution and orange juice are studied, and the result shows that in two solution there are many difference in fluorescence spectra of potassium sorbate, but the fluorescence characteristic peak exists in ,λex/λem =375/490 nm. It can be seen from the two dimensional fluorescence spectra that the relationship between the fluorescence intensity and the concentration of potassium sorbate is very complex, so there is no linear relationship between them. To determine the concentration of potassium sorbate in orange juice, a new method combining Particle Swarm Optimization （PSO） algorithm with Back Propagation （BP） neural network is proposed. The relative error of two predicted concentrations is 1.83% and 1.53% respectively, which indicate that the method is feasible. The PSO-BP neural network can accurately measure the concentration of potassium sorbate in orange juice in the range of 0.1-2.0g·L^-1.

关 键 词：荧光光谱 山梨酸钾 PSO-BP神经网络 浓度测定 

分 类 号：O433.1[机械工程—光学工程]