广东化工2024,Vol.51Issue(16) :116-118,144.DOI:10.3969/j.issn.1007-1865.2024.016.037

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

郝玉良 马时霖 马成详 苏然 陈爱军 刘浩 李秋蓉
广东化工2024,Vol.51Issue(16) :116-118,144.DOI:10.3969/j.issn.1007-1865.2024.016.037

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

郝玉良 1马时霖 2马成详 3苏然 1陈爱军 1刘浩 1李秋蓉1
扫码查看

作者信息

  • 1. 昆仑数智科技有限责任公司,北京 102206
  • 2. 中国石油天然气股份有限公司兰州石化分公司,甘肃 兰州 730060
  • 3. 中石油兰州石化榆林化工有限公司,陕西 榆林 719199
  • 折叠

摘要

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.

Abstract

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.

关键词

深度学习/混合建模/实时优化/模拟计算/操作优化

Key words

deep learning/hybrid modeling/real-time optimization/simulation calculation/operation optimization

引用本文复制引用

出版年

2024
广东化工
广东省石油化工研究院

广东化工

影响因子:0.288
ISSN:1007-1865
段落导航相关论文