PhD Thesis Proposal: Neal Seegmiller
Formulation and Calibration of Fast, Accurate Vehicle Motion Models
Carnegie Mellon University
November 30, 2012, 10:00 a.m., NSH 3002
High performance wheeled mobile robots (WMRs) require fast, accurate motion models for planning and control. This thesis addresses two challenges to producing these models: the tradeoff between fidelity and speed in model formulation, and the need for laborious calibration procedures. To address the first challenge, I propose the formulation of “enhanced” 3D velocity kinematic models that provide comparable accuracy to comprehensive physics-based models, but at a fraction of the computational cost. In a modular way for any WMR design, the motion prediction problem is formulated as the solution of a differential-algebraic equation (DAE), with wheel velocity constraints derived using a vector algebra-based approach. These models account for articulations on uneven terrain, and are enhanced to account for slip, powertrain dynamics, and extreme conditions such as rollover.
To address the second challenge, I propose the convenient self-calibration of these models using an online integrated equation error (IEE) approach. The IEE approach also enables the characterization of non-systematic error, such that the planner may reason about the uncertainty of motion predictions. I will prove experimentally that models formulated and calibrated as proposed are both faster and more accurate than alternatives. This thesis unifies and extends successful prior work by Alonzo Kelly, Forrest Rogers-Marcovitz, and myself.
Alonzo Kelly, Chair
Karl Iagnemma, Massachusetts Institute of Technology