基于SFLA和LSSVM的芹菜总黄酮含量的预测模型  

Predicting Content of Total Flavonoids in Celery Based on SFLA and LSSVM

作　　者：曾燕[1] 成新文[1] 王晓[1] 陈欲云[2] 

机构地区：[1]四川理工学院计算机学院,四川自贡643000 [2]四川理工学院化学与制药工程学院,四川自贡643000

出　　处：《湖北农业科学》2015年第6期1486-1489,共4页Hubei Agricultural Sciences

摘　　要：利用混合蛙跳算法(SFLA)对最小二乘支持向量机(LSSVM)的参数进行优化,建立了一种基于混合蛙跳算法和最小二乘支持向量机的芹菜总黄酮提取预测模型。仿真结果表明,该预测模型提高了预测精度,性能更加稳定,为芹菜总黄酮提取的在线预估和优化控制提供了可靠的技术参考。Shuffled frog leaping algorithm （SFLA） was used to parameters of optimize least square support vector machine （LSSVM）. The prediction model of extracting total flavonoids from celery based on shuffled frog leaping algorithm and least squares support vector machine was established. Simulation results showed that this prediction method had high accuracy and stable performance. It was conducive to achieve online estimation and optimal control.

关 键 词：芹菜 混合蛙跳算法 最小二乘支持向量机 总黄酮 预测 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]