Date of Award

8-2014

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Electrical Engineering

Advisor

Burg, Timothy C.

Committee Member

Groff, Richard E.

Committee Member

Walker, Ian D.

Committee Member

Wagner, John R.

Abstract

An innovative type of mobile manipulator, designated Manipulator on VTOL (Vertical Take-Off and Landing) Aircraft (MOVA), is proposed as a potential candidate for autonomous execution of field work in less-structured indoor and outdoor environments. Practical use of the MOVA system requires a unified controller that addresses the coupled and complex dynamics of the composite system; especially the interaction of the robotic manipulator with the aircraft airframe. Model-based controller design methods require explicit dynamics models of the MOVA system. Preliminary investigation of a two-dimensional MOVA system toward a dynamics model and controller design is presented in preparation for developing the controller of the more complex MOVA system in 3D space. Dynamics of the planar MOVA system are derived using the Lagrangian approach and then transforming the result into a form that facilitates controller design using the concept of a virtual manipulator. A MOVA end-effector trajectory tracking controller was designed with the transformed dynamics equation using the integrator back-stepping control design framework. Validity of the controller is shown via stability analysis, simulation results, and results from a physical test-bed. A systematic approach is illustrated for the derivation of the 3D MOVA system dynamics equations. The resulting dynamics equations are represented abstractly in the standard robot dynamics form and proven to have the skew-symmetric property, which is a useful property for control derivation. An open source Mathematica program was developed to achieve automatic symbolic derivation of the MOVA system dynamics. Accessory tools were also designed to create a tool-chain that starts with an Autodesk Inventor CAD drawing, generates input to the Mathematica program, and then formats the output for direct use in MATLAB and Simulink. A unified nonlinear control algorithm that controls the 3D MOVA system, including both the aircraft and the onboard manipulator, as a single entity was developed to achieve trajectory tracking of the MOVA end-effector position and attitude based on the explicit dynamics equation. Globally Uniformly Ultimately Bounded (GUUB) stability is proven for the controller using Lyapunov-type stability analysis. Physical testing was constructed in order to to demonstrate the performance of the proposed controller on a MOVA system with a two-link onboard manipulator.

Share

COinS