- 5
- 0
- 约2.53万字
- 约 4页
- 2019-01-14 发布于天津
- 举报
基于PSO和LSSVM的生化过程建模研究
专题研究
文章编号:1001-9944(2009)02-0005-04
基于PSO 和LSSVM 的生化过程建模研究
吴 洲,潘 丰,田 鹏
(江南大学 通信与控制工程学院,无锡 214122)
摘要:针对最小二乘支持向量机(LSSVM )在生化过程建模中的重要建模参数值选择问题,
提出利用具有较强的全局搜索能力的粒子群(PSO )优化算法,对最小二乘支持向量机建模
过程中的重要参数进行优化调整, 每一个粒子的位置向量对应一组最小二乘支持向量机
建模的参数。 利用参数优化调整后得到的具有较优拟合预测效果的模型对谷氨酸发酵过
程进行预测,仿真结果表明,该方法能使模型取得较好的预测效果。
关键词:最小二乘支持向量机;粒子群;参数选择
中图分类号: 文献标志码:
TP391.9 A
Modeling of Biochemical Process Based on Particle Swarm Optimization and
Least Squares Support Vector Machine
, ,
WU Zhou PAN Feng TIAN Peng
( , , , )
School of Communication and Control Engineering Jiangnan University Wuxi 214122 China
: ’
Abstract In order to overcome the disadvantage that it s difficult to get better parameter value in biochemical mod-
,
eling of least squares support vector machine a method is put forward to finding the better parameter value by parti-
,
cle swarm optimization which has the better search ability. Each particle indicates a group of LSSVM parameters and
the particles will change unceasingly until satisfy the condition. The proposed method is applied to establish a soft-
sensor model for Glutanic acid fermentation process .And carries on the forecast using this model. The simulation re-
sult indicated that the method can cause the model to obtain the better for
原创力文档

文档评论(0)