A surrogate-based optimization framework for simultaneous synthesis of chemical process and heat exchanger network

作     者：Li, Mingxin Yu Zhuang Li, Weida Dong, Yachao Lei Zhang Jian Du Shen Shengqiang 

作者机构：Dalian Univ Technol Sch Chem Engn Inst Proc Syst Engn Dalian 116024 Liaoning Peoples R China Dalian Univ Technol Sch Energy & Power Engn Key Lab Liaoning Prov Desalinat Dalian 116024 Liaoning Peoples R China Nucl Power Inst China Res Inst 4 Chengdu 610200 Sichuan Peoples R China 

出 版 物：《CHEMICAL ENGINEERING RESEARCH & DESIGN》 (化学工程师协会汇刊，A辑：化学工程研究与设计)

年 卷 期：2021年第170卷

页      面：180-188页

核心收录：

学科分类：0817[工学-化学工程与技术] 08[工学] 

基　　金：National Natural Science Foundation of China [22008023, 21776035] China Postdoctoral Science Foundation [2019TQ0045] 

主　　题：Surrogate model Simultaneous optimization Process synthesis Heat exchanger network Mathematical programming 

摘      要：Heat-integrated process synthesis is fundamental to achieve higher energy efficiency. The well-known sequential-conceptual methods have been widely adopted to solve the synthesis problem in a hierarchical manner. However, the natural hierarchy fails to consider complex interactions between the unit operation and the heat integration. To address this issue, a surrogate-based optimization framework is proposed for simultaneous synthesis of chemical process and heat exchanger network. An artificial neural network (ANN)-based surrogate model, derived from the simulation data generated via rigorous mechanism modelling approach, is established for process units to replace their complex realistic models. With surrogate model formulation incorporated into heat integration, an enhanced transshipment-based mixed integer nonlinear programming model is introduced to synthesize heat exchanger network with variable flowrates and temperatures, aiming at the maximized annual profit. Finally, two example studies are investigated to demonstrate the effectiveness of the proposed framework. (c) 2021 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.