TY - JOUR
T1 - Autonomous Navigation of Robots
T2 - Optimization with DQN
AU - Escobar-Naranjo, Juan
AU - Caiza, Gustavo
AU - Ayala, Paulina
AU - Jordan, Edisson
AU - Garcia, Carlos A.
AU - Garcia, Marcelo V.
N1 - Publisher Copyright:
© 2023 by the authors.
PY - 2023/6
Y1 - 2023/6
N2 - Featured Application: The application of “Autonomous Navigation of Robots: Optimization with DQN” involves using reinforcement learning techniques to optimize the navigation of autonomous robots. Specifically, the Deep Q-Network (DQN) algorithm is used to train a robot to make decisions based on sensory inputs and to learn optimal paths to reach a goal. This can have a wide range of potential applications, such as in manufacturing and logistics, where robots can autonomously navigate and transport materials within a warehouse; or in search and rescue missions, where robots can navigate through disaster-stricken areas to locate and rescue survivors. By optimizing the navigation process with DQN, these robots can operate more efficiently and safely in their respective environments. In the field of artificial intelligence, control systems for mobile robots have undergone significant advancements, particularly within the realm of autonomous learning. However, previous studies have primarily focused on predefined paths, neglecting real-time obstacle avoidance and trajectory reconfiguration. This research introduces a novel algorithm that integrates reinforcement learning with the Deep Q-Network (DQN) to empower an agent with the ability to execute actions, gather information from a simulated environment in Gazebo, and maximize rewards. Through a series of carefully designed experiments, the algorithm’s parameters were meticulously configured, and its performance was rigorously validated. Unlike conventional navigation systems, our approach embraces the exploration of the environment, facilitating effective trajectory planning based on acquired knowledge. By leveraging randomized training conditions within a simulated environment, the DQN network exhibits superior capabilities in computing complex functions compared to traditional methods. This breakthrough underscores the potential of our algorithm to significantly enhance the autonomous learning capacities of mobile robots.
AB - Featured Application: The application of “Autonomous Navigation of Robots: Optimization with DQN” involves using reinforcement learning techniques to optimize the navigation of autonomous robots. Specifically, the Deep Q-Network (DQN) algorithm is used to train a robot to make decisions based on sensory inputs and to learn optimal paths to reach a goal. This can have a wide range of potential applications, such as in manufacturing and logistics, where robots can autonomously navigate and transport materials within a warehouse; or in search and rescue missions, where robots can navigate through disaster-stricken areas to locate and rescue survivors. By optimizing the navigation process with DQN, these robots can operate more efficiently and safely in their respective environments. In the field of artificial intelligence, control systems for mobile robots have undergone significant advancements, particularly within the realm of autonomous learning. However, previous studies have primarily focused on predefined paths, neglecting real-time obstacle avoidance and trajectory reconfiguration. This research introduces a novel algorithm that integrates reinforcement learning with the Deep Q-Network (DQN) to empower an agent with the ability to execute actions, gather information from a simulated environment in Gazebo, and maximize rewards. Through a series of carefully designed experiments, the algorithm’s parameters were meticulously configured, and its performance was rigorously validated. Unlike conventional navigation systems, our approach embraces the exploration of the environment, facilitating effective trajectory planning based on acquired knowledge. By leveraging randomized training conditions within a simulated environment, the DQN network exhibits superior capabilities in computing complex functions compared to traditional methods. This breakthrough underscores the potential of our algorithm to significantly enhance the autonomous learning capacities of mobile robots.
KW - artificial intelligence
KW - autonomous navigation
KW - optimization DQN (Deep Q-Network)
KW - robotics
UR - http://www.scopus.com/inward/record.url?scp=85163975009&partnerID=8YFLogxK
U2 - 10.3390/app13127202
DO - 10.3390/app13127202
M3 - Article
AN - SCOPUS:85163975009
SN - 2076-3417
VL - 13
JO - Applied Sciences (Switzerland)
JF - Applied Sciences (Switzerland)
IS - 12
M1 - 7202
ER -