The NEOShield-2 project, within which this thesis was built, aims to develop technologies for mitigations measures, in order to be able to deviate a NEO forecast to impact Earth from its initial trajectory. In order to assess the success of a NEO orbit deflection, a reconnaissance spacecraft is used. This thesis focuses on the estimation of the relative state between the NEO and this observer spacecraft, based on image processing techniques. From a series of mono-camera images, the relative pose between the NEO and the spacecraft can be obtained. A Kalman filter is then implemented, in order to increase the relative state estimation accuracy, and to estimate the inertial state of the NEO. On the first hand, it allows to estimate some state variables which are not directly measured. On the second hand, it removes part of the noise which is contained in the image processing measurements. Because the image processing algorithms are sensitive to the environment, the measurement noise is not Gaussian and outliers are contained in the measurements. This work compared different techniques to handle these outliers as well as the non-Gaussian character of the noise. A consistency analysis is first provided in order to select only the techniques which ensure the consistency of the filter. Second, the filter performance is compared for the techniques which make the filter consistent. This thesis concludes that the overall consistency and performance of the filter can be improved with respect to its original version.
-
Relative navigation at close proximity of small solar system bodies
HARBULOT, QUENTIN, LOUIS, JULES
2015/2016
Abstract
The NEOShield-2 project, within which this thesis was built, aims to develop technologies for mitigations measures, in order to be able to deviate a NEO forecast to impact Earth from its initial trajectory. In order to assess the success of a NEO orbit deflection, a reconnaissance spacecraft is used. This thesis focuses on the estimation of the relative state between the NEO and this observer spacecraft, based on image processing techniques. From a series of mono-camera images, the relative pose between the NEO and the spacecraft can be obtained. A Kalman filter is then implemented, in order to increase the relative state estimation accuracy, and to estimate the inertial state of the NEO. On the first hand, it allows to estimate some state variables which are not directly measured. On the second hand, it removes part of the noise which is contained in the image processing measurements. Because the image processing algorithms are sensitive to the environment, the measurement noise is not Gaussian and outliers are contained in the measurements. This work compared different techniques to handle these outliers as well as the non-Gaussian character of the noise. A consistency analysis is first provided in order to select only the techniques which ensure the consistency of the filter. Second, the filter performance is compared for the techniques which make the filter consistent. This thesis concludes that the overall consistency and performance of the filter can be improved with respect to its original version.File | Dimensione | Formato | |
---|---|---|---|
Thesis.pdf
non accessibile
Descrizione: Thesis text
Dimensione
72.81 MB
Formato
Adobe PDF
|
72.81 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/134065