Two-Stage Soft-Sensing Model Based On PLS And LSSVM Algorithm

As a new and expanding measurement technique, soft-sensing technique is mainly applied to estimate the parameters which are hard to be measured or cannot be measured on-line with existing instruments.the production process of industry is complicated,nonlinear, and time-delay, the existing methods can't solve all the problems simultaneously. To solve this problem, a new modeling method based on partial least squares method and least squares support vector machine method is researched in this article.This paper researches partial least squares and support vector machines modeling methods deeply,following the main contents:(1) The theory and methods of soft-sensing are summaried all sidedly. The advantages and limitations of the methods are researched deeply.(2) The partial least squares method and least squares support vector machine method are mainly researched in this chapter. And the limitation of the two methods are summarized through the analysis of the theory and the synthetic test.(3) Two-stage soft-sensing method based on partial least squares method and least squares support vector machine method is applied in establishing the model of the melt index. Through comparing with the outcome of LSSVM, The effectiveness of two-stage method is verified...