Speaker: Justin Pearson (Electrical and Computer Engineering)
Title: The Kalman Filter: An unreasonably good state estimator
Date: Nov 24, 2014
Location: 1601 Elings Hall, University of California, Santa Barbara
The Kalman Filter stirred up excitement in the aerospace industry in the 60's as an intuitive, optimal, and easily-implementable state estimator. It allowed control engineers to employ celebrated state-feedback controllers even when the system's state was not directly measurable, in a plug-and-play sort of way. This excitement spread to other industries, who discovered that the filter works unreasonably well even when violating the assumptions underlying the design of the filter, like having a non-Gaussian noise distribution. Several extensions of the KF apply the same basic concepts to nonlinear systems (the extended / unscented Kalman Filter) and statistically-simulated systems (the particle filter). In this talk, I will explain the basics of control theory and linear systems, describe the Kalman Filter, and show some cool examples.
... Sorry about the occasionally fuzzy audio!
More videos, projects, code, presentations, contact info: http://justinppearson.com
(c) 2014 Justin P. Pearson. All rights reserved.
