Automated synthesis of steady-state continuous processes using reinforcement learning

作     者：Quirin Göttl Dominik GGrimm Jakob Burger Quirin G(o)ttl;Dominik G.Grimm;Jakob Burger

作者机构：Technical University of MunichCampus Straubing for Biotechnology and SustainabilityLaboratory of Chemical Process Engineering94315 StraubingGermany Technical University of MunichCampus Straubing for Biotechnology and Sustainability94315 StraubingGermany Weihenstephan-Triesdorf University of Applied Sciences94315 StraubingGermany Technical University of MunichDepartment of Informatics85748 GarchingGermany 

出 版 物：《Frontiers of Chemical Science and Engineering》 (化学科学与工程前沿（英文版）)

年 卷 期：2022年第16卷第2期

页      面：288-302页

核心收录：

学科分类：081704[工学-应用化学] 07[理学] 08[工学] 0817[工学-化学工程与技术] 070303[理学-有机化学] 0703[理学-化学] 

基　　金：Projekt DEAL 

主　　题：automated process synthesis flowsheet synthesis artificial intelligence machine learning reinforcement learning 

摘      要：Automated flowsheet synthesis is an important field in computer-aided process *** present work demonstrates how reinforcement learning can be used for automated flowsheet synthesis without any heuristics or prior knowledge of conceptual *** environment consists of a steady-state flowsheet simulator that contains all physical *** agent is trained to take discrete actions and sequentially build up flowsheets that solve a given process problem.A novel method named SynGameZero is developed to ensure good exploration schemes in the complex ***,flowsheet synthesis is modelled as a game of two competing *** agent plays this game against itself during training and consists of an artificial neural network and a tree search for forward *** method is applied successfully to a reaction-distillation process in a quaternary system.