State estimation for constrained systems with redundant coordinates
Conference Paper
Overview
Identity
View All
Overview
abstract
Classically, a minimal coordinate representation of a dynamical system is found before applying a Kalman filter to estimate the system states. For many applications, it is more desirable to instead use a redundant coordinate representation of the system due to physical geometry, control design, or location of sensors and control actuators. However, a redundant coordinate description necessitates the introduction of holonomic constraints to augment the governing differential equations for an accurate model of the system dynamics. In this paper, a new approach for applying the Continuous-Discrete Extended Kalman Filter to dynamical systems with redundant coordinates is presented as an alternative to the traditional minimal coordinate process. A two-armed robotic manipulator is presented as a relevant, complex example for which implementation of this estimation technique for redundant coordinates is advantageous. 2011 by Julie J. Parish and John E. Hurtado.
