论文部分内容阅读
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。
For the ethylene distillation process, this paper proposes an online optimization structure model. In order to avoid the time delay between the complex distillation process modeling and the interface program with the large-scale flow simulation software and to improve the speed of simulation, this model combined with the learning function of neural network is used to simulate the distillation process. The improved genetic algorithm was used to realize the online optimization system of the debutanizer in the ethylene digital factory. The algorithm introduced the pattern search technology and acceleration strategy to meet the time and accuracy requirements of online optimization. The results show that this method has the advantages of short operation time and fast convergence, and can meet the requirements of online optimization. At the same time, this optimization model can be extended to other processes.