文献类型：期刊文献

中文题名：基于特征选择的常减压装置模型及在计划优化中的应用

英文题名：Modeling of crude oil distillation units based on feature selection and its application in planning optimization

作者：李勇[1];钱锋[1];宋育梅[2]

机构：[1]华东理工大学化工过程先进控制和优化技术教育部重点实验室,上海200237;[2]中国石化上海石油化工股份有限公司,上海200540

年份：2021

卷号：72

期号：3

起止页码：1419

中文期刊名：化工学报

外文期刊名：CIESC Journal

收录：CSTPCD;;EI(收录号：20211510212221);Scopus;北大核心:【北大核心2020】；CSCD:【CSCD2021_2022】；

基金：国家自然科学基金基础科学中心项目(61988101);重点国际(地区)合作研究项目(61720106008);国家杰出青年科学基金项目(61725301);中央高校基本科研业务费专项资金(222201917006)。

语种：中文

中文关键词：常减压装置;方程组模型;特征选择;鲸鱼优化算法;计划优化

外文关键词：crude oil distillation units;equations model;feature selection;whale optimization algorithm;plan optimization

摘要：常减压装置将原油切割为不同中间产品,其作为炼油工艺的龙头装置,对炼油过程生产计划排产与效益提升至关重要。通过建立一种高精度且具有良好求解效率的常减压模型,用以求解模型关键指标实沸点(TBP)曲线,即综合考虑切割产品的实沸点(TBP)与原油TBP、流量、温度等变量影响,构建非线性方程组模型来表征输入输出间的关系;利用特征选择方法遴选相关变量(包括进料性质、相邻TBP及其二次项等),并采用鲸鱼优化算法优化方程组系数。仿真结果表明,该多输出相互关系模型与已有文献工作相比,在兼顾求解效率基础上,常减压装置各蒸馏切割产品TBP曲线预测上有更高的精度,将此模型应用到炼厂计划优化中,与传统的悬摆切割模型对比,优化结果优于传统悬摆切割模型。
Crude oil distillation units separate crude oil into different intermediate products.As a leading device in the refining process,it is essential for the production planning and efficiency improvement of the refining process.The purpose of this article is to establish a crude oil distillation units model with high-precision and good solution efficiency.In order to solve the key indicators of true boiling point(TBP)curve,We comprehensively consider the influence of the product????s TBP and crude oil TBP,flow rate,temperature and other variables,and construct a nonlinear equation model to characterize the relationship between input and output.We also use the feature selection method to select the relevant variables(including feed properties,adjacent TBP and its quadratic terms,etc.),and choose the whale optimization algorithm to optimize the coefficients of the equations.The simulation results show that the model has higher accuracy in predicting the TBP curve of each distillation and cutting product of crude oil distillation units compared with the existing literature work, and this model is applied to refineryplanning optimization. Compared with the traditional pendulum cutting model, the optimization result is better thanthe traditional swing-cut model.