摘要
在偏最小二乘回归和样条变换理论研究的基础上,提出炼油装置常压塔航油干点的软测量。采用偏最小二乘回归方法筛选一种辅助变量和建立航油干点的软测量模型。仿真结果表明,本方法选择的辅助变量携带信息量大,对主导变量解释能力强。如样本集相同,比RBF网络和支持向量机软测量模型预测精度高,泛化能力强。
On the basis of a theoretical research of partial least-squares regression spline transformation,the soft sensoring of aviation kerosene dry point of atmospheric distilation column of oil refining device was proposed.According to the collected samples,by using the auxiliary variables choosing method and the soft sensing method proposed,a soft sensoring model for predicting the dry point of aviation kerosene was set up.The simulation result shows that auxiliary variables chosen out by the proposed method car...
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2009年第3期300-304,共5页
Computers and Applied Chemistry
关键词
偏最小二乘回归
样条变换
软测量
干点
partial least-squares regression
spline transformation
soft sensoring
dry point