文献类型：期刊文献

英文题名：A General Methodology for Kinetic Parameter Estimation of Industrial Propylene Multistage Polymerization Process Using High-Efficient Enhanced RIME Algorithm

作者：Zhang, Xixiang[1];Lu, Jingyi[1,2];Tian, Zhou[1,2];Du, Wenli[1,2];Qian, Feng[1]

机构：[1]East China Univ Sci & Technol, Key Lab Smart Mfg Energy Chem Proc, Minist Educ, Shanghai 200237, Peoples R China;[2]Qingyuan Innovat Lab, Quanzhou 362801, Fujian, Peoples R China

年份：2025

卷号：64

期号：33

起止页码：16404

外文期刊名：INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH

收录：;EI(收录号：20253619099097);WOS:【SCI-EXPANDED(收录号:WOS:001545724000001)】；

基金：The authors gratefully acknowledge the financial support of this work by the National Key R&D Program of China (2022YFB3304903).

语种：英文

外文关键词：Chains - Copolymerization - Ethylene - Evolutionary algorithms - Industrial plants - Kinetic parameters - Optimization - Parameter estimation - Random processes - Sensitivity analysis

摘要：Accurate estimation of kinetic parameters in propylene (PP) multistage polymerization process is crucial for modeling and optimization of industrial plants. These parameters significantly influence the predictions of the chain microstructures and macroscopic properties of polymer. This work aims to propose a generic methodology for estimating kinetic parameters of propylene homo- and copolymerization processes, allowing one to accurately predict the macro-behavior (i.e., production rate) and the properties (i.e., melt flow index, molecular weight distribution (MWD), ethylene content, copolymer composition distribution (CCD), and comonomer sequence length distribution (CSLD)) of polymer in industrial polypropylene process. A novel enhanced RIME algorithm (ERIME), incorporating selective evolutionary crossover and Gaussian random walk mechanisms, is proposed for the parameter estimation framework. Initially, a microstructure-oriented process model is developed, comprising a kinetic model, a nonideal reactor model, and chain microstructure models. Simultaneous deconvolution of MWDs from both reactors is carried out to determine the number of active site types, as well as the mass fraction and chain microstructures of polymer at each site type. Subsequently, the parameters are screened through parameter reduction and sensitivity analysis to identify the subset for estimate. The kinetic parameter estimation problem is then formulated as an optimization problem, minimizing the least-squares difference between industrial production targets and model predictions, and is solved using the ERIME algorithm efficiently. The process model, based on the estimated kinetic parameters, has been validated by simulation-based and industrial cases. These results confirm the applicability, potential, and efficiency of the systematic methodology.