Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


36467.pdf (10.81 MB)

ETD Abstract Container

Abstract Header

Modular Architecture for Intelligent Aerial Manipulators

Medhi, Jishu K
http://rave.ohiolink.edu/etdc/view?acc_num=ucin1573811910421278

Abstract Details

2019, MS, University of Cincinnati, Engineering and Applied Science: Aerospace Engineering.
Unmanned aerial vehicles are becoming increasingly popular all over the world, with the development of the technology allowing it to be shifted out of the research lab and into the commercial sector. For quite a while now, UAVs have been used by various institutions for military surveillance, aerial photography, and agricultural crop monitoring. Although most commercially available civilian drones do an excellent job at surveillance tasks, these platforms generally are not equipped to handle tasks that require direct environmental intervention and interaction. Although humans are usually better suited to freely interact with low-altitude surroundings (within 100 feet of the ground), there are several instances where direct human interaction may not be possible because the environment may be too dangerous, hostile, or inaccessible to people. In situations like these, the unique capabilities of an aerial manipulator – an unmanned aerial vehicle that has a robotic manipulator mechanically attached to its body to manipulate its environment – can come into the picture. This thesis presents research work done to implement and test a modular and resilient system architecture for aerial manipulator operations. The system architecture used in this research is loosely based on NASA JPL's Resilient Spacecraft Executive, which was initially intended to robustly handle uncertainties for spacecraft operations. The entire system architecture of the aerial manipulator is comprised of three different layers – deliberative, habitual and reflexive – each of which communicate back and forth with each other via a common message passing framework, such as the Robot Operating System (ROS) (which was used in this implementation). A major part of this research involved the use of a mid-fidelity dynamic simulation environment with a physics model available that could accurately show `good’ and `mis’behavior of the aerial manipulator in offline testing, prior to tests involving real physical hardware. As such, the ROS-native Gazebo 3D simulator and MoveIt interface were used for simulation and path planning of the manipulator, respectively. To validate the feasibility of the proposed system architecture, test cases for three different phases of flight with four manipulator actions are considered, and the results of the simulation are presented. Note that the source code and models created and modified for use in this work are all available for general use under a BSD 3-clause open source license at https://github.com/medhijk/ROS_quadrotor_simulator.
Catharine McGhan, Ph.D. (Committee Chair)
Ou Ma, Ph.D. (Committee Member)
Rajnikant Sharma, Ph.D. (Committee Member)
155 p.
Aerospace Materials

Recommended Citations

Citations

  • Medhi, J. K. (2019). Modular Architecture for Intelligent Aerial Manipulators [Master's thesis, University of Cincinnati]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1573811910421278

    APA Style (7th edition)

  • Medhi, Jishu. Modular Architecture for Intelligent Aerial Manipulators. 2019. University of Cincinnati, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=ucin1573811910421278.

    MLA Style (8th edition)

  • Medhi, Jishu. "Modular Architecture for Intelligent Aerial Manipulators." Master's thesis, University of Cincinnati, 2019. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1573811910421278

    Chicago Manual of Style (17th edition)

ucin1573811910421278
213
© 2019, all rights reserved.
This open access ETD is published by University of Cincinnati and OhioLINK.