摘要
介绍了一种自适应优化方法，通过机理分析和模型结构辨识确定系统的模型结构，采用在线运行数据拟合过程动态模型参数，并依此计算目标函数相对于各操作参数的梯度，最终确定优化的方向．根据算法设计的计算机软件ＡＮＯＰＴ在一个丙烯腈装置上得到实际应用．结果表明，本方法具有适应性好，抗干扰能力强，寻优速度快等特点，并且不需在线组分分析仪表，非常适用于实际工业过程．
This paper proposes an adaptive optimization method. The model structure is selected through mechanism analysis and model structure identification. The dynamic model parameters are identified on line using the process data, then the gradients of the object function for every operating variable are computed based on the model identified. Finally the optimal direction to move the working point is decided. The software ANOPT designed based on the algorithm has been applied to an industrial acrylonitrile apparatus and the result shows that this method has good adaptability, high anti interference and quick optimization speed. As the method doesn′t need any on line component analyzer, it is suitable for actual industrial process.
出处
《自动化学报》
EI
CSCD
北大核心
1998年第1期73-80,共8页
Acta Automatica Sinica
关键词
系统辨识
自适应优化法
丙烯腈装置
System identification, steady state optimization, fluid bed reactor, acrylonitrile.
