{"title":"Deep reinforcement learning for process design: Review and perspective","authors":"Qinghe Gao, Artur M Schweidtmann","doi":"10.1016/j.coche.2024.101012","DOIUrl":null,"url":null,"abstract":"<div><p>The transformation toward renewable energy and feedstock supply in the chemical industry requires new conceptual process design approaches. Recently, deep reinforcement learning (RL), a subclass of machine learning, has shown the potential to solve complex decision-making problems and aid sustainable process design. However, its suitability in static process design still needs to be examined. We discuss the advantages and disadvantages of RL for process design. Then, we survey state-of-the-art research through three major elements: (1) information representation, (2) agent architecture, and (3) environment and reward. Moreover, we discuss perspectives on underlying challenges and promising future works to unfold the full potential of RL for process design in chemical engineering.</p></div>","PeriodicalId":292,"journal":{"name":"Current Opinion in Chemical Engineering","volume":"44 ","pages":"Article 101012"},"PeriodicalIF":8.0000,"publicationDate":"2024-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2211339824000133/pdfft?md5=48434807719d11339aadf8ef46d44883&pid=1-s2.0-S2211339824000133-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Current Opinion in Chemical Engineering","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2211339824000133","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"BIOTECHNOLOGY & APPLIED MICROBIOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
The transformation toward renewable energy and feedstock supply in the chemical industry requires new conceptual process design approaches. Recently, deep reinforcement learning (RL), a subclass of machine learning, has shown the potential to solve complex decision-making problems and aid sustainable process design. However, its suitability in static process design still needs to be examined. We discuss the advantages and disadvantages of RL for process design. Then, we survey state-of-the-art research through three major elements: (1) information representation, (2) agent architecture, and (3) environment and reward. Moreover, we discuss perspectives on underlying challenges and promising future works to unfold the full potential of RL for process design in chemical engineering.
期刊介绍:
Current Opinion in Chemical Engineering is devoted to bringing forth short and focused review articles written by experts on current advances in different areas of chemical engineering. Only invited review articles will be published.
The goals of each review article in Current Opinion in Chemical Engineering are:
1. To acquaint the reader/researcher with the most important recent papers in the given topic.
2. To provide the reader with the views/opinions of the expert in each topic.
The reviews are short (about 2500 words or 5-10 printed pages with figures) and serve as an invaluable source of information for researchers, teachers, professionals and students. The reviews also aim to stimulate exchange of ideas among experts.
Themed sections:
Each review will focus on particular aspects of one of the following themed sections of chemical engineering:
1. Nanotechnology
2. Energy and environmental engineering
3. Biotechnology and bioprocess engineering
4. Biological engineering (covering tissue engineering, regenerative medicine, drug delivery)
5. Separation engineering (covering membrane technologies, adsorbents, desalination, distillation etc.)
6. Materials engineering (covering biomaterials, inorganic especially ceramic materials, nanostructured materials).
7. Process systems engineering
8. Reaction engineering and catalysis.