{"title":"Deep Reinforcement Learning-Based Optimization Framework with Continuous Action Space for LNG Liquefaction Processes","authors":"Jieun Lee, Kyungtae Park","doi":"10.1007/s11814-025-00428-x","DOIUrl":null,"url":null,"abstract":"<div><p>Recently, the application of reinforcement learning in process systems engineering has attracted significant attention recently. However, the optimization of chemical processes using this approach faces various challenges related to performance and stability. This paper presents a process optimization framework using a continuous advantage actor–critic that is modified from the existing advantage actor–critic algorithm by incorporating a normal distribution for action sampling in a continuous space. The proposed reinforcement learning-based optimization framework was found to outperform the conventional method in optimizing a single mixed refrigerant process with 10 variables, achieving a lower specific energy consumption value of 0.294 kWh/kg compared to the value of 0.307 kWh/kg obtained using the genetic algorithm. Parametric studies performed into the hyperparameters of the continuous advantage actor-critic algorithm, including the maximum episodes, learning rate, maximum action value, and structures of the neural networks, are presented to investigate their impacts on the optimization performance. The optimal specific energy consumption, namely 0.287 kWh/kg, was achieved by varying the learning rate from the base case to 0.00005. These results demonstrate that reinforcement learning can be effectively applied to the optimization of chemical processes.</p></div>","PeriodicalId":684,"journal":{"name":"Korean Journal of Chemical Engineering","volume":"42 8","pages":"1613 - 1628"},"PeriodicalIF":3.2000,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Korean Journal of Chemical Engineering","FirstCategoryId":"5","ListUrlMain":"https://link.springer.com/article/10.1007/s11814-025-00428-x","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"CHEMISTRY, MULTIDISCIPLINARY","Score":null,"Total":0}
引用次数: 0
Abstract
Recently, the application of reinforcement learning in process systems engineering has attracted significant attention recently. However, the optimization of chemical processes using this approach faces various challenges related to performance and stability. This paper presents a process optimization framework using a continuous advantage actor–critic that is modified from the existing advantage actor–critic algorithm by incorporating a normal distribution for action sampling in a continuous space. The proposed reinforcement learning-based optimization framework was found to outperform the conventional method in optimizing a single mixed refrigerant process with 10 variables, achieving a lower specific energy consumption value of 0.294 kWh/kg compared to the value of 0.307 kWh/kg obtained using the genetic algorithm. Parametric studies performed into the hyperparameters of the continuous advantage actor-critic algorithm, including the maximum episodes, learning rate, maximum action value, and structures of the neural networks, are presented to investigate their impacts on the optimization performance. The optimal specific energy consumption, namely 0.287 kWh/kg, was achieved by varying the learning rate from the base case to 0.00005. These results demonstrate that reinforcement learning can be effectively applied to the optimization of chemical processes.
期刊介绍:
The Korean Journal of Chemical Engineering provides a global forum for the dissemination of research in chemical engineering. The Journal publishes significant research results obtained in the Asia-Pacific region, and simultaneously introduces recent technical progress made in other areas of the world to this region. Submitted research papers must be of potential industrial significance and specifically concerned with chemical engineering. The editors will give preference to papers having a clearly stated practical scope and applicability in the areas of chemical engineering, and to those where new theoretical concepts are supported by new experimental details. The Journal also regularly publishes featured reviews on emerging and industrially important subjects of chemical engineering as well as selected papers presented at international conferences on the subjects.