基于混合模型法的精馏塔热耦合技术研究

Study on thermal-coupled distillation columns technology based on hybrid model method

针对非理想物系,提出了基于混合模型法的精馏塔热耦合参数优化方法,充分结合了Aspen软件对过程的准确模拟和神经元网络简洁计算的优点,首先获得参数寻优的可行域,然后利用遗传算法对目标函数进行全局寻优。将优化后的热耦合技术应用于中国石油吉林分公司脱乙烷塔与乙烯精馏塔,可节省能耗22.9%。该方法还可用于相关工业过程热耦合技术中。

In this paper,a thermal-coupled distillation columns method for optimizing parameter is presented based on a hybrid model method,which fully uses the advantages of Aspen Plus' accurate simulation and neural network's brief calculation.The feasible range is obtained from the precise model simulation of Aspen Plus;the correlation between input and output variables is then modeled by neural network;and the optimization is done together with genetic algorithm.This method after optimized is used in the dethanizer and ethylene ditillation tower in Jinlin Chemical Corp.,and it can save energy consumption by 22.9%.And this method can also be used in other thermal coupled processes.