基于机理与数据融合的乙烯智能优化研究

浏览(5) 下载(0)

关键词:: 乙烯装置；机理建模；数据驱动；双模融合；操作优化
ethylene plant; mechanism modeling; data-driven model; hybrid modeling; operation optimization

摘要:: 针对乙烯装置操作调控依赖人工经验、质量控制滞后、数据价值未充分挖掘等问题，提

出 1 种基于机理与数据双模融合的智能操作优化软件方案。采用 Coilsim 软件搭建裂解炉机理

模型，结合深度学习与遗传算法构建数据驱动模型，开发了在线模拟与优化平台。实际应用表

明，该软件可实现乙烯纯度由 99.939 6%提至 99.949 2%，丙烯产量由 39.53 t/h 提高至 40.91 t/h，

验证了双模融合方法的有效性和经济性。

Aiming at the problems of manual-dependent operation control, time-lagged quality control, and underutilized data
value in ethylene plants, this paper proposes an intelligent operation optimization software solution based on hybrid mechanismdata modeling. A mechanism model of cracking furnaces is built using Coilsim software, combined with deep learning and genetic
algorithm to construct a data-driven model, and an online simulation and optimization platform is developed. Practical application
shows that the software can increase ethylene purity from 99.939 6% to 99.949 2% and propylene production from 39.53 t/h to
40.91 t/h, verifying the effectiveness and economic benefits of the hybrid modeling approach.