This chapter discusses a class of widely used models derived from simple equations of motion—constant velocity and constant acceleration. The discrete‐time kinematic model obtained by discretizing the continuous‐time state space representation driven by white noise is presented. The state model defined directly in discrete time using a piecewise constant white random sequence as process noise is presented. Explicit filters for noiseless kinematic models are explained and explicit steady‐state filters for noisy kinematic models are derived. The statistical characterization of the process noise is a key filter design parameter, which is discussed in detail. Subsequently, one important aspect of tracking system design—selection of the sampling frequency, i.e., the revisit rate—is discussed. The models are presented for a single coordinate, and some intuitive insight into filter design is presented. A problem solving section appears at the end of the chapter.
Estimation for Kinematic Models
04.01.2002
33 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Springer Verlag | 2023
|Kinematic State Estimation for a Mars Rover
NTRS | 1999
|Kinematic state estimation for a Mars rover
Tema Archiv | 2000
|