Use of Deep Learning Methodologies in Combination with Reinforcement Techniques within Autonomous Mobile Cyber-physical Systems
Volodymyr Levytskyi, Pavlo Kruk, Oleksii Lopuha, Danylo Sereda, Viktor Sapaiev, Oleksii Matsiievskyi · 2024
This study proposes a solution to the problem that arises during the development and operation of cyber-physical systems (CPS) based on the capabilities of artificial intelligence (AI). The study presents a comprehensive overview of CPS and their interactions with the environment. It offers a general classification of deep reinforcement learning algorithms, followed by an in-depth analysis of each algorithm. The current utilization of deep learning technologies within autonomous mobile cyber-physical systems is discussed. It proposes various approaches to integrate deep learning technologies with reinforcement learning in autonomous mobile CPS. Additionally, it provides a generalized structural diagram illustrating the incorporation of a deep learning module with reinforcement into CPS. The result of the research is the use of deep learning technology to improve the performance of autonomous mobile cyber-physical systems.