非线性递推部分最小二乘及其应用

部分最小二乘回归(PLS)可较好地解决变量间的共线性问题,目前被广泛地应用于过程建模和监控.本文将递推PLS(RPLS)算法同RBF网络相结合,给出了一种非线性递推PLS方法(NRPLS),可根据在线数据自适应地调整模型结构和参数,使模型适应非线性过程的变化.在确定RBF网络的隐层节点参数时,采用了一种改进的k-means聚类算法,自动确定最优的聚类区数.该递推算法用于聚丙稀熔融指数软测量模型的在线修正,取得了较好的效果.

Nonlinear Recursive Partial Least Squares and Its Applications

Partial least squares (PLS) regression is widely used in process modeling and monitoring to deal with a large number of variables with collinearity. In this paper, a nonlinear recursive PLS (NRPLS) algorithm is proposed by combining recursive PLS(RPLS) with RBF network. It can modify model structure and parameters to adapt process changes according to new data. For determining the Gaussian centers in RBF network, an improved k-means clustering algorithm is developed to find the optimal clusters automatically. An application of the technology to soft measurement of polypropylene melt index is demonstrated. The experimental results show the effectiveness of this algorithm.