Deep reinforcement learning for feedback control in a collective flashing ratchet
A collective flashing ratchet transports Brownian particles using a spatially periodic, asymmetric, and time-dependent on-off switchable potential. The net current of the particles in this system can be substantially increased by feedback control based on the particle positions. Several feedback pol...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
American Physical Society
2021-04-01
|
| Series: | Physical Review Research |
| Online Access: | http://doi.org/10.1103/PhysRevResearch.3.L022002 |
| _version_ | 1797211057582768128 |
|---|---|
| author | Dong-Kyum Kim Hawoong Jeong |
| author_facet | Dong-Kyum Kim Hawoong Jeong |
| author_sort | Dong-Kyum Kim |
| collection | DOAJ |
| description | A collective flashing ratchet transports Brownian particles using a spatially periodic, asymmetric, and time-dependent on-off switchable potential. The net current of the particles in this system can be substantially increased by feedback control based on the particle positions. Several feedback policies for maximizing the current have been proposed, but optimal policies have not been found for a moderate number of particles. Here, we use deep reinforcement learning (RL) to find optimal policies, with results showing that policies built with a suitable neural network architecture outperform the previous policies. Moreover, even in a time-delayed feedback situation where the on-off switching of the potential is delayed, we demonstrate that the policies provided by deep RL provide higher currents than the previous strategies. |
| first_indexed | 2024-04-24T10:20:26Z |
| format | Article |
| id | doaj.art-5ed5687dda4e40f98f42efed1969ba15 |
| institution | Directory Open Access Journal |
| issn | 2643-1564 |
| language | English |
| last_indexed | 2024-04-24T10:20:26Z |
| publishDate | 2021-04-01 |
| publisher | American Physical Society |
| record_format | Article |
| series | Physical Review Research |
| spelling | doaj.art-5ed5687dda4e40f98f42efed1969ba152024-04-12T17:08:53ZengAmerican Physical SocietyPhysical Review Research2643-15642021-04-0132L02200210.1103/PhysRevResearch.3.L022002Deep reinforcement learning for feedback control in a collective flashing ratchetDong-Kyum KimHawoong JeongA collective flashing ratchet transports Brownian particles using a spatially periodic, asymmetric, and time-dependent on-off switchable potential. The net current of the particles in this system can be substantially increased by feedback control based on the particle positions. Several feedback policies for maximizing the current have been proposed, but optimal policies have not been found for a moderate number of particles. Here, we use deep reinforcement learning (RL) to find optimal policies, with results showing that policies built with a suitable neural network architecture outperform the previous policies. Moreover, even in a time-delayed feedback situation where the on-off switching of the potential is delayed, we demonstrate that the policies provided by deep RL provide higher currents than the previous strategies.http://doi.org/10.1103/PhysRevResearch.3.L022002 |
| spellingShingle | Dong-Kyum Kim Hawoong Jeong Deep reinforcement learning for feedback control in a collective flashing ratchet Physical Review Research |
| title | Deep reinforcement learning for feedback control in a collective flashing ratchet |
| title_full | Deep reinforcement learning for feedback control in a collective flashing ratchet |
| title_fullStr | Deep reinforcement learning for feedback control in a collective flashing ratchet |
| title_full_unstemmed | Deep reinforcement learning for feedback control in a collective flashing ratchet |
| title_short | Deep reinforcement learning for feedback control in a collective flashing ratchet |
| title_sort | deep reinforcement learning for feedback control in a collective flashing ratchet |
| url | http://doi.org/10.1103/PhysRevResearch.3.L022002 |
| work_keys_str_mv | AT dongkyumkim deepreinforcementlearningforfeedbackcontrolinacollectiveflashingratchet AT hawoongjeong deepreinforcementlearningforfeedbackcontrolinacollectiveflashingratchet |