Thesis (MScEng)--Stellenbosch University, 2011. / ENGLISH ABSTRACT: This thesis presents the development of a state estimation system for use in an Autonomous
Airborne Refueling (AAR) operation through the simulated implementation of GPS, monocular
and stereoscopic vision, inertial measurement sensors and boom parameter measurement in
combination with the Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF).
A set of functional criteria for the estimation system was developed through an analysis of
the control system input requirements and associated constraints. The estimation system
is further developed by integrating the sensor configurations into the estimation algorithm
structures through the derivation of the applicable mathematical models. Final sensor configurations
are set based on a sensitivity analysis in which the effect of parameters such as
sensor noise, placement and quantity are related to the accuracy with which the states are
estimated.
Uncertainty in the process noise, which is typically approximated, is overcome by adding an
adaptive element to the estimation algorithms in which the current process noise is estimated
allowing compensation for unmodeled process noise uncertainty.
Finally twelve practical sensor configurations are established utilising unique combinations
of the five sensors. Each configuration is simulated using both estimation algorithms after
which all results are evaluated with respect to one another as well as to the minimum state
accuracy criteria. Conclusions are presented based on the evaluation of the results followed
by recommendation for future development. / AFRIKAANSE OPSOMMINGS: Die ontwikkeling van ’n toestandafskattingstelsel, spesifiek toegepas op outonome brandstofhervulling,
word voorgelê in hierdie tesis. Hierdie ontwikkeling behels die implementering
van GPS, monukulêre- en stereo-visie sensors, inersiële sensor eenhede en verbindingsarmsensors
wat gebruik word in ’n Uitgebruide Kalman Filter (Extended Kalman Filter) en Geurlose
Kalman Filter (Unscented Kalman Filter).
’n Volledige ontleding van die beheerstelsel se toevoervereistes en geassosieerde beperkings
is gebruik om ’n stel beoordelingsmaatstawwe vir die toestandafskatting-stelsel te bepaal.
Die stelsel is verder ontwikkel deur verskillende sensorkonfigurasies met die afskattingsalgoritmes
te kombineer deur die afleiding van toepaslike wiskundinge modelle. Hierdie konfigurasies
is verfyn deur ’n sensitiwiteitsanalise, waar die verwantskap tussen die effekte van
sensorruis, sensorligging, hoeveelheid sensors ondersoek is met betrekking tot afskattingsakkuraatheid.
Onsekerheid in die stelsel se prosesruis is deur ’n aanpassings substelsel hanteer, wat kompensasie
vir ongemodeleerde onsekerheid moontlik maak. Twaalf praktiese sensorkonfigurasies
is opgestel vanuit unieke kombinasies van die vyf sensore behartig in die projek. Hierdie
konfigurasies is deur beide afskattingsalgoritmes gebruik om sodoende die akkuraatheid van
die konfigurasies asook die afskattingsalgoritmes te evalueer met betrekking tot mekaar en
aan die hand van die beoordelingsmaatstawwe vir die beheerstelsel. Die tesis is afgesluit deur
gevolgtrekkings asook aanbevelings vir toekomstige navorsing.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:sun/oai:scholar.sun.ac.za:10019.1/17852 |
| Date | 12 1900 |
| Creators | Runhaar, Anton Johan |
| Contributors | Peddle, I. K., Stellenbosch University. Faculty of Engineering. Dept. of Electrical and Electronic Engineering. |
| Publisher | Stellenbosch : Stellenbosch University |
| Source Sets | South African National ETD Portal |
| Language | en_ZA |
| Detected Language | Unknown |
| Type | Thesis |
| Format | 155 p. : ill. |
| Rights | Stellenbosch University |
Page generated in 0.0018 seconds