【作者】 王振雷； 钱锋；

【Author】 Wang Zhenlei Qian Feng2 State-Key Laboratory of Chemical Engineering , Eastern China Univ. of Science and Technology, Shanghai 200237, China

【机构】 华东理工大学化学工程联合国家重点实验室；

【摘要】 本文针对一种典型化工装置——丙烯精馏塔的工艺特征和操作特点,提出一种基于支持向量机(SVM)软测量技术,对塔釜丙烯浓度进行"测量",进而设计了丙烯浓度智能控制系统,对塔釜丙烯指标进行实时控制,解决了塔釜丙烯浓度超标且不能在线测量的问题。现场应用效果表明,该系统具有良好的测量精度,取得了良好的控制效果,满足了工业现场运行的需要。更多还原

【Abstract】 The propylene concentration in the bottom of the propylene distillation column is the key quality index which need be controlled. In some plants, the bottom propylene concentration is determined only by off-line laboratory. The data can not be used online. A kind of control structure of the propylene concentration is developed. Integrating the principle model of the column and the site variables data, we propose the soft sensor of the propylene concentration based on the principal component analysis (PCA) and support vector machines (SVM). The intelligent inference control system is developed basing on the soft sensor. And the control strategy is applied on the propylene distillation column in one olefin plant. The control results are demonstrated in the paper, too.更多还原