- Title
- On the energy and geometry of vehicle guidance, navigation and control
- Creator
- Renton, Chris
- Relation
- University of Newcastle Research Higher Degree Thesis
- Resource Type
- thesis
- Date
- 2014
- Description
- Research Doctorate - Doctor of Philosophy (PhD)
- Description
- The thesis revisits the intimate connection between the geometry of the configuration space of vehicles and the descriptions of their energy and power and exploits it to provide novel solutions to particular problems in motion control and navigation. Starting from the geometry of the configuration space, one can build additional related spaces. Elements of these spaces, together with those in the configuration space, can then be used to describe scalar storage and content functions, which relate to energy and power respectively. Once these spaces are constructed and the scalar functions adopted, geometric properties can be exploited for designing guidance and motion control strategies. The choice of coordinates assigned to each of the related spaces have bearing on the difficulty of the description of the dynamics and the design of control strategies. For example, it is well known that the choice of configuration coordinates may be used to eliminate holonomic constraints while the choice of momentum coordinates may be used to eliminate nonholonomic constraints. In cases where these coordinate choices can simplify the problem, they also favour the use of a particular storage function. The same geometric tools and choice of coordinates used to approach vehicle motion control may be used to address vision-based navigation. Indeed, the image obtained from cameras mounted on the vehicle can be projected onto a space of particular geometry, which is akin to a configuration space. From the latter, one can again build additional spaces that may be exploited to describe the phenomena of optic flow and other image features and use this to build measurement models to be used for estimation. This thesis is separated into two parts. The first part revisits the connections between energy and geometry to obtain a rigid-body model of vehicle dynamics. We exploit the symmetry highlighted explicitly in the bond-graph formalism to relate the Euler-Lagrange (EL), Hamiltonian, and Brayton-Moser (BM) equations by proposing what is believed to be a new set of unifying model equations. The proposed model retains the advantage of the BM model in that the dissipation and interconnection ports are included ab initio in the model---features that the EL and Hamiltonian models must be extended to support. This part of the thesis then considers a class of mechanical systems, which includes vehicles, and addresses the problem of position set-point regulation using port-Hamiltonian theory. The geometry of the configuration space is used to construct the basis for solutions of the matching partial differential equations that allows manifold regulation for underactuated vehicles. We present examples of underactuated mechanical systems that illustrate the application of the proposed control design method. These examples include a marine surface vessel and a quadrotor helicopter. We also discuss examples beyond vehicles such as a robot manipulator mounted on a flexible foundation. The second part of the thesis considers the problem of visual navigation. A new model for optic flow is developed for use in state estimation. This model is motivated by a particular choice of state, which is inspired by the geometric aspects of the problem. This choice of state is also exploited to outline adaptations to visual simultaneous localisation and mapping (SLAM), including the fusion with optic flow, and the tracking of landmarks that exploits the observable directions in the optic flow field.
- Subject
- vehicle guidance; motion control; visual navigation; unifying model equations
- Identifier
- http://hdl.handle.net/1959.13/1048142
- Identifier
- uon:14879
- Rights
- Copyright 2014 Chris Renton
- Language
- eng
- Full Text
- Hits: 406
- Visitors: 791
- Downloads: 330
| Thumbnail | File | Description | Size | Format | |||
|---|---|---|---|---|---|---|---|
| View Details Download | ATTACHMENT01 | Abstract | 354 KB | Adobe Acrobat PDF | View Details Download | ||
| View Details Download | ATTACHMENT02 | Thesis | 3 MB | Adobe Acrobat PDF | View Details Download |