文献类型：会议论文

英文题名：Octane model based on principal component analysis and neural network by near-infrared spectroscopy and its application in online gasoline blending process

作者：He, Kaixun[1]; Cheng, Hui[1]; Qian, Feng[1]

机构：[1] Key Laboratory of Advanced Control and Optimization for Chemical Processes, Ministry of Education, East China University of Science and Technology, Shanghai, 200237, China

会议日期：October 18, 2013 - October 21, 2013

会议地点：No. 130, Meilong Road, Shanghai, China

语种：英文

外文关键词：Infrared devices - Neural networks - Spectrum analysis - Gasoline - Pipelines - Near infrared spectroscopy - Blending - Least squares approximations

摘要：For the gasoline pipeline blending process, recipe optimization system is greatly dependent on the near-infrared spectroscopy online analyzer, whose spectral model plays an important role in the measurement. The sepectral model's accuracy and adaptability directly affect the applicability of the entire online blending system. This paper studies how to establish model for gasoline octane number for the gasoline pipeline blending process with near-infrared spectroscopy online analyzer. It is proposed using principal component analysis (PCA) together with Artificial Neural Network (ANN) method to establish spectral-model for octane number. Multivariate linear regressions(MLR) and partial least squares (PLS) method have also been used to establish gasoline octane model for comparison purpose. The results show that the model established by PCA and ANN has strong anti-jamming capability and suitable for gasoline online blending application.