Kanarski, Christian, Kaulen, Bastian, Kühne, Frederik, Wisch, Tim Owe, Gussow, Karoline and Christensen, Sören (2023) Deep Reinforcement Learning for Autonomous SONAR Port Monitoring. In: Fortschritte der Akustik - DAGA 2023. , ed. by von Estorff, Otto and Lippert, Stephan. Deutsche Gesellschaft für Akustik e.V. (DEGA), Berlin, pp. 196-199. ISBN 978-3-939296-21-8

Preview

Text kanarski_et_al_daga_2023.pdf Download (634kB) | Preview

Abstract

The use of MIMO-SONAR systems to autonomously monitor a port environment requires a robust control of the system parametrization and its adaptation to changing environmental conditions in real-time. Deep reinforcement learning (DRL) can be used to implement a control-assisting artificial intelligence (AI) which adapts the system parametrization in relation to the observed environment scans. Through the design of a reward function, the controlling agent can learn the fulfillment of given sub-goals, such as the evaluation of security risks and the management of limited computational and energy resources. During training, the agent explores the unknown environmental dynamics in a trial-and-error fashion and improves its policy by exploiting the gathered experiences of agent-environment interactions. By retrospective analysis of the chosen system parametrization and the resulting scan observations, the agent learns to adapt its monitoring strategy to fulfill the main goal of reliably detecting unwanted intruders inside of the port. This work presents the design of the reward function, the training architecture, and the performance evaluation of the trained deep reinforcement learning agent for a simulated port environment.

Actions (login required)

View Item