【摘要】 精馏塔是化工过程中最常用的操作单元 ,具有很强的非线性和时变性 ,故很难进行机理建模分析或常规在线实时控制 ,因而提出一种基于径向基函数神经网络的优化控制方案。通过利用径向基函数神经网络建立精馏塔产品质量的软测量模型 ,将软测量结果与现场数据比较 ,表明本模型具有比较准确的跟踪显示效果 ,并将软测量模型进一步应用到精馏塔的回流量和釜液排放量的优化控制中更多还原

【Abstract】 Distillation columns are the most usual operating units in the process of chemical engineering. Their non-linear and time-varying characteristics made the design of the control scheme very difficult. The paper proposed an optimizing control of distillation columns based on RBF neural networks,which found product estimation model of distillation columns through RBF neural networks. The results of product estimation were proved by the data collected from productive process, besides, which were applied to advanced optimizing control of the circulation & tower emission of distillation process.更多还原