Return to search

DEVELOPMENT OF PASSIVE VISION BASED RELATIVE STATION KEEPING FOR UNMANNED SURFACE VEHICLES

<p dir="ltr">Unmanned surface vehicles (USVs) offer a versatile platform for various maritime applications, including research, surveillance, and search-and-rescue operations. A critical capability for USVs is maintaining position (station keeping) in dynamic environments and coordinating movement with other USVs (formation control) for collaborative missions. This thesis investigates control strategies for USVs operating in challenging conditions. </p><p dir="ltr">The initial focus is on evaluating traditional control methods like Backstepping and Sliding Mode controllers for station keeping in simulated environments with disturbances. The results from these tests pointed towards the need for a more robust control technique, like deep-learning based control for enhanced performance. </p><p dir="ltr">The thesis then explores formation control, a crucial aspect of cooperative USV missions. A vision-based passive control strategy utilizing a virtual leader concept is proposed. This approach leverages onboard cameras to detect markers on other USVs, eliminating the need for direct communication and potentially improving scalability and resilience. </p><p dir="ltr">Then the thesis presents vision-based formation control architecture and the station keeping controller evaluations. Simulation results are presented, analyzed, and used to draw conclusions about the effectiveness of the proposed approaches. Finally, the thesis discusses the implications of the findings and proposes potential future research directions</p>

  1. 10.25394/pgs.25675221.v1
Identiferoai:union.ndltd.org:purdue.edu/oai:figshare.com:article/25675221
Date26 April 2024
CreatorsAjinkya Avinash Chaudhary (18430029)
Source SetsPurdue University
Detected LanguageEnglish
TypeText, Thesis
RightsCC BY 4.0
Relationhttps://figshare.com/articles/thesis/DEVELOPMENT_OF_PASSIVE_VISION_BASED_RELATIVE_STATION_KEEPING_FOR_UNMANNED_SURFACE_VEHICLES/25675221

Page generated in 0.0017 seconds