Generating collective foraging behavior for robotic swarm using deep reinforcement learning
- PDF / 1,691,392 Bytes
- 8 Pages / 595.276 x 790.866 pts Page_size
- 9 Downloads / 190 Views
ORIGINAL ARTICLE
Generating collective foraging behavior for robotic swarm using deep reinforcement learning Boyin Jin1 · Yupeng Liang1 · Ziyao Han1 · Kazuhiro Ohkura1 Received: 20 May 2020 / Accepted: 14 September 2020 / Published online: 19 October 2020 © International Society of Artificial Life and Robotics (ISAROB) 2020
Abstract This paper mainly discussed the generation of collective behaviors with raw camera images as the primary information input. The swarm robotic system exhibits considerable advantages when faced with individual-level failure or the lack of global information. Spatial information has always been a necessity in generating collective transport behavior. The rise of deep neural network technology makes it possible for a robot to perceive the environment from its visual input. In this paper, the use of deep reinforcement learning in training a robotic swarm to generate collective foraging behavior is shown. The collective foraging behavior is evaluated in a transportation task, where robots need to learn to process image information while cooperatively transport foods to the nest. We applied a deep Q-Learning algorithm and several improved versions to develop controllers for robotic swarms. The results of computer simulations show that using images as the main information input can successfully generate collective foraging behavior. Besides, we also combine the advantages of several algorithms to improve performance and perform experiments to examine the flexibility of the developed controllers. Keywords Swarm robotics · Visual information · Deep reinforcement learning · Deep Q-Learning
1 Introduction Swarm robotics (SR) [1] is the study of how a large number of relatively simple physically embodied agents can be designed, such that a desired collective behavior emerges from the local interactions among the agents and between the agent and the environment. Agents in the SR systems are only able to perceive limited information from the environment. They have to work cooperatively to accomplish the task which cannot achieve by a single agent. To achieve a transportation behavior in a swarm or a single robot, situational awareness is necessary; however, global position information is neither effective nor practical. This work was presented in part at the 3rd International Symposium on Swarm Behavior and Bio-Inspired Robotics (Okinawa, Japan, November 20–22, 2019). * Kazuhiro Ohkura kohkura@hiroshima‑u.ac.jp Boyin Jin [email protected]‑u.ac.jp 1
Hiroshima University, 1‑4‑1, Kagamiyama, Higashi‑Hiroshima, Hiroshima 739‑8527, Japan
13
Vol:.(1234567890)
In nature, social animals have sensory feedback such as vision [2], auditory sensation, or tactile sensation under limited local sensory capabilities. Creatures including swarm perceived environment via high-dimensional information [3] composed of all kinds of sensory information input. The difficulty of using high-dimensional information inputs like camera images will significantly enlarge the amount of data that need to be processed by the
Data Loading...
