遗传算法在非线性原油调合中的应用研究
Application of Genetic Algorithm for non-linear Crude Oil Blending
张锐1，周家庆2，陈夕松1，亚宏宇3
作者单位：（1.东南大学自动化学院 江苏 南京210096；2.南京富岛信息工程有限公司 江苏 南京210061；3.中国石油化工股份有限公司金陵分公司生产计划处，江苏 南京210033）
中文关键字：原油调合；非线性；遗传算法
英文关键字：Crude oil blending; non-linear; Genetic algorithm
中文摘要：原油在线调合优化技术近年来已开始在炼化企业得到成功应用，目前主要以线性优化技术解决原油关键指标（如硫含量、酸含量等）的调合问题。然而，原油调合还存在非线性指标的调合，论文提出了一种面向原油非线性指标控制的原油在线调合优化方法，建立非线性原油调合模型，采用遗传算法进行求解，实例验证了该方法的实用性和有效性。
英文摘要：On line crude oil blending has been successfully applied in refineries in recent years, and linear optimization technologies have usually been employed for the key crude oil properties blending, such as sulfur, acid, etc. However, non-linear properties widely exist in crude oil blending. This paper proposes a non-linear crude oil on-line blending method in which a non-linear crude oil blending model is constructed first, and genetic algorithm is put forward to access the final solution. The real industrial practice has proved its practicality and effectiveness.