Guidance navigation and control solutions for unmanned heterogeneous vehicles during a collaborative mission

Abstract : The theme of "low-cost navigation", characterized by transferring performance from sensors to data fusion and control algorithms, is a central theme for many military applications, in particular those related to light troops. Using inertial or magnetic sensors from the MEMS category, along with civilian GPS, optimizes weight, size and energy consumption. As will be demonstrated, this approach requires application-specific guidance navigation and control algorithms to compensate for the relatively poor quality of the sensor signals. In this thesis, we consider three scenarios of interest and develop innovative techniques for guidance and navigation. First, we consider the case of a ground vehicle equipped with proximity sensors, a gyroscope, odometers and a GPS. Experimentally, we implement an algorithm for obstacle avoidance for which we establish a proof of convergence, and an offline path-planning algorithm that is complemented by a dynamic feedback controller and a non-linear state estimator. Then, we investigate the case of an unstable helicopter. We develop and implement algorithms within a real-time control system that we designed specifically for computational and estimation tasks. The state estimator includes a model of the flight dynamics, and uses data from inertial sensors, a barometer, and a GPS to serve as input for a closed-loop controller. The parameters of the model are identified accurately by using data obtained during experimental flights. Eventually, we shall perform autonomous hovering flights, which stress the performance of the system. Finally, we consider the problem of a pedestrian walking inside buildings. The heading estimation errors observed using various platforms (both ground and aerial) suggest a new use of the magnetic field, and we propose to derive information by inspecting its gradients. We explain how to use (unknown) disturbances of the magnetic field to improve the position estimate in inertial navigation. Experimental results emphasize the relevance of the approach.
Document type :
Theses
Complete list of metadatas

Cited literature [88 references]  Display  Hide  Download

https://pastel.archives-ouvertes.fr/pastel-00004492
Contributor : Ecole Mines Paristech <>
Submitted on : Monday, December 22, 2008 - 8:00:00 AM
Last modification on : Monday, November 12, 2018 - 10:57:02 AM
Long-term archiving on: Friday, September 10, 2010 - 1:12:29 PM

Identifiers

  • HAL Id : pastel-00004492, version 1

Citation

David Vissière. Guidance navigation and control solutions for unmanned heterogeneous vehicles during a collaborative mission. Mathematics [math]. École Nationale Supérieure des Mines de Paris, 2008. English. ⟨NNT : 2008ENMP1540⟩. ⟨pastel-00004492⟩

Share

Metrics

Record views

1945

Files downloads

2667