Automation of membrane capacitive deionization process using reinforcement learning. (1st December 2022)

• Record Type: Journal Article
• Authors: Yoon, Nakyung; Park, Sanghun; Son, Moon; Cho, Kyung Hwa
• Is Part Of: Water research. Volume 227(2022)
• Publication Date: 2022-12-01
• Subjects: Membrane capacitive deionization -- Deep reinforcement learning -- Automation -- Optimization; Water -- Pollution -- Research -- Periodicals; 363.7394
• DOI: 10.1016/j.watres.2022.119337
• Languages: English
• ISSNs: 0043-1354

Abstract

Highlights: Deep reinforcement learning (DRL) was applied to automate MCDI process. The numerical MCDI transport model is combined as the environment. The DRL model reduces energy consumption for desalination by 22%. Abstract: Capacitive deionization (CDI) is an alternative desalination technology that uses electrochemical ion separation. Although several attempts have been made to maximize the energy efficiency and productivity of CDI with conventional control methods, it is difficult to optimize the CDI processes because of the complex correlation between the operational conditions and the composition of feed water. To address these challenges, we applied deep reinforcement learning (DRL) to automatically control the membrane capacitive deionization (MCDI) process, which is one of the representative CDI processes, to accomplish high energy efficiency while desalinating water. In the DRL model, the numerical model is combined as the environment that provides states according to the actions. The feed water conditions, that is, the input state of the DRL, were assumed to have a random salt concentration and constant foulant concentration. The model was constructed to minimize energy consumption and maximize desalted water volume per cycle. After training of 1, 000 episodes, the DRL model achieved a 22.07% reduction in specific energy consumption (from 0.054 to 0.042 kWh m −3 ) and 11.60% increase in water desalted water volume per cycle (from 1.96×10 −5 to 2.19×10 −5 m 3 ), achieving the desired degree of desalination, compared to the first episode. This improved performance was because the trained model selected the optimized operating conditions of current, voltage, and the number and intensity of flushing. Furthermore, it was possible to train the model depending on demand by modifying the reward function of the DRL model. The fundamental principle described in this study for applying the DRL model in MCDI operations can be the cornerstone of a fully automated water desalination process.