UKF-SLAM Implementation for the Optical Navigation System of a Lunar Lander

No Thumbnail Available

URL

Journal Title

Journal ISSN

Volume Title

Sähkötekniikan korkeakoulu | Master's thesis
Ask about the availability of the thesis by sending email to the Aalto University Learning Centre oppimiskeskus@aalto.fi

Date

2017-10-23

Department

Major/Subject

Space Robotics and Automation

Mcode

AS3004

Degree programme

Erasmus Mundus Space Master

Language

en

Pages

59+6

Series

Abstract

This thesis presents a state estimator for the optical navigation system of an autonomous lunar lander. The proposed estimator addresses the Simultaneous Localization and Mapping problem, SLAM, combined with the Unscented Kalman Filter. The ultimate goal of the thesis is to verify whether the state estimation can rely on a relative positioning scheme when measurements of the position of the vehicle, in world coordinates, are not available. Therefore, an image feature tracking routine is proposed as a reference for localization. In order to handle this functionality and guarantee reliable state estimation any time, the estimator is founded on a monocular SLAM approach, jointly with the Inverse Depth Parametrization. This work describes in detail the UKF-SLAM routine with the corresponding state transition model and observation model. In the context of SLAM, a novel approach for the initialization of newly observed image features in the state vector and covariance matrix, based on the Unscented Transform, is introduced to the tracking routine. Subsequently, the development method and validation tests are presented. Finally, the results of the simulations and final conclusions are discussed. The results acknowledge the capability of the UKFSLAM scheme as a suitable estimator for the aforementioned navigation system, in the absence or in presence of absolute position measurements.

Description

Supervisor

Visala, Arto

Thesis advisor

Ammann, Nikolaus
Enmark, Anita

Keywords

unscented Kalman filter, simultaneous localization and mapping, inverse depth parametrization, pinhole camera model

Other note

Citation