文献类型：期刊文献

英文题名：Fast and Effective Dynamic Optimization for Chemical Processes with Catalyst Deactivation Based on Incremental Encoding and Random Search

作者：Guo, Jingjing[1];Du, Wenli[1];Wu, Qun[1];Ye, Zhencheng[1]

机构：[1]East China Univ Sci & Technol, Key Lab Adv Control & Optimizat Chem Proc, Minist Educ, Shanghai 200237, Peoples R China

年份：2021

卷号：60

期号：7

起止页码：2983

外文期刊名：INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH

收录：;EI(收录号：20210909983798);WOS:【SCI-EXPANDED(收录号:WOS:000623232700021)】；

基金：This work was supported by the National Natural Science Foundation of China (Basic Science Center Program: 61988101), the International (Regional) Cooperation and Exchange Project (61720106008), the National Natural Science Fund for Distinguished Young Scholars (61725301), the National Natural Science Foundation of China (Major Program: 61590923), and the Fundamental Research Funds for the Central Universities (222202017006).

语种：英文

外文关键词：Catalyst deactivation - Catalyst activity - Signal encoding - Dynamic programming - Heuristic algorithms - Encoding (symbols) - Iterative methods

摘要：Dynamic is widely encountered in the chemical industry, which may come from equipment aging or catalyst deactivation. Meanwhile, the drift of these variables generally has a certain trend of change, such as worse equipment status or lower catalyst activity. Most of the existing optimization methods focus on steady-state optimization problems, and the optimization schemes proposed for such dynamic optimization problems are generally time-consuming. Therefore, to achieve optimal economic benefits, drifting operating conditions are challenges that must be overcome. In this work, to use the prior knowledge of the research object and guide the algorithm to converge fast, incremental encoding (IE) combined with the population is introduced, which searches for a feasible and better control trajectory. Next, the coarse random search (RS) method often used in iterative dynamic programming is introduced to improve the performance of the algorithm. The proposed two-step IE-RS optimization algorithm based on the control vector parameterization (CVP) combines the advantages of a heuristic algorithm and iterative dynamic optimization, which not only ensures fast convergence but also ensures the effectiveness of the algorithm, and is finally demonstrated in the acetylene hydrogenation process.