基于机理和数据混合建模的炼化装置操作优化研究及应用

Optimizing Software for Intelligent Operation of Refining and Chemical Plants

随着制造行业智能化的发展,模拟计算大量应用于化工生产过程。本文利用混合建模技术设计了一套炼化装置智能操作优化软件,采用深度学习算法和遗传算法对模型进行优化。该软件能够实时预测产品质量、优化操作参数、进行数据预警。相比于传统的机理建模速度更快准确度更高,提高了炼厂操作过程的精细化程度、运行过程的稳定性以及生产过程的产品质量。

With the development of intelligent manufacturing industry,simulation calculation is widely used in chemical production process.This article used hybrid modeling technology to design a set of intelligent operation optimization software for refining and chemical equipment,and used deep learning algorithms and genetic algorithms to optimize the model.This software could predict product quality in real-time,optimize operating parameters,and provide data alerts.Compared to traditional mechanism modeling,it was faster and more accurate,could improve the refinement of the refinery operation process,the stability of the operation process,and the product quality of the production process.