Using Probability Estimates to Identify Environmental Features for a Nonholonomic Control System
| Autor: | Steven B. Skaar, Humberto Arriola, John-David Yoder |
|---|---|
| Rok vydání: | 1997 |
| Předmět: |
Nonholonomic system
Orientation (computer vision) Covariance matrix Computer science Applied Mathematics Process (computing) Aerospace Engineering Kalman filter Extended Kalman filter Nonlinear system Space and Planetary Science Control and Systems Engineering Control theory Position (vector) Electrical and Electronic Engineering Algorithm |
| Zdroj: | Journal of Guidance, Control, and Dynamics. 20:1215-1220 |
| ISSN: | 1533-3884 0731-5090 |
| DOI: | 10.2514/2.4179 |
| Popis: | A third-order set of nonlinear, ordinary differential equations models the relationship between internally measurablewheel rotationsand the position and orientation of an automatically guided vehicle, buttheserelationships areimprecise, growing increasingly inadequateastheirintegrals, and thevehicle,proceedfrom pointofdeparture. An extended Kalman e lter (EKF) is used to combine video observations of features on that portion of the environment that does not move, together with the sensed wheel rotations, to produce the ongoing estimates needed for navigation. The experimental usefulness is examined of a byproduct of the e lter, the estimate error covariance matrix, to an integrally related process: the process of identifying video observations with features of known location within the environment; these identities are required for application of new vision observations to the stateestimates. Thegoodnessof theEKF’ sprobability density functions isexperimentally examined by comparing them against actual, accumulated data; experimental results are presented from the use of an extensive theoretical developmentthatassesses, basedonrelativeprobabilitiesinferredfrom thesedistributions, theidentitiesofdensely occurring, nondistinct cues. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|