Support Vector Machine-based Soft Sensors in the Isomerisation Process

This paper presents the development of soft sensor empirical models using support<br /> vector machine (SVM) for continual assessment 2,3-dimethylbutane and 2-methylpentane mole percentage as important product quality indicators in refinery isomerisation process. During model development, critical steps were taken, including selection pre-processing industrial process data, which are broadly discussed this paper. The SVM results compared with dynamic linear output error nonlinear Hammerstein-Wiener model. Evaluation developed on independent data sets showed their reliability component contents. sensors to be embedded into control system, serve primarily a replacement during analysersb failure service periods.