作　　者： ; ; ; ;

机构地区： 华南理工大学化学与化工学院

出　　处： 《计算机与应用化学》 2009年第6期794-798,共5页

摘　　要： 针对乙烯精馏过程,本文提出在线优化结构模型。为了避免复杂精馏过程建模和与大型流程模拟软件接口程序的时间延迟,提高模拟运算速度,该模型结合神经网络的学习功能,用于精馏过程的模拟计算。采用改进的遗传算法,实现乙烯数字工厂中脱丁烷塔的在线优化系统,算法引入模式搜索技术和加速策略,满足在线优化的时间和精度要求。计算结果表明,该法具有运算时间短、收敛快的优点,能够满足在线优化的要求。同时,为该优化模型拓展到其它工艺过程提供参考。 Taking ethylene distillation process as background, the optimization model of online system is developed in this text. The learning ability of artificial neural network is carried to realize the steady state simulation of distillation process. That advances the operating speed of simulation, and avoids modeling from complex distillation process and consuming a lot of time on the interface to the process simulation software. Improved Genetic Algorithms （IGA） realizes the online optimization system of the debutanizer in the digital ethylene plant. Pattern search method and acceleration strategy are introduced to meet the requirements of online optimization in the time and accuracy. The results show that this method has the advantages of short run time and fast convergence, which can better meet the requirements of online optimization, and provides a basis for expanding to other processes.

关 键 词： 神经网络 遗传算法 模式搜索 加速策略

领　　域： [自动化与计算机技术] [自动化与计算机技术] [化学工程]