Font Size:
Path Planning and Decision Making for Co-operative Multiple UAVs
Brian White
Last modified: 2010-05-09
Abstract
The use of multiple UAVs for mapping and surveillance is becoming practical in the near future. As such, autonomous behaviour that is robust and predictable is necessary for such use. This presentation will describe the role of decision making and path planning in producing an autonomous system that can survey areas that have some rural areas and some urban areas, together with some structured waterways. This implies that paths need to follow route structures that are related to road networks, waterways or other natural features that require specific paths, such as valleys, forest clearings.
In order to produce an autonomous system that has specific guaranteed behaviours, the link between the decision making code and the path planning needs to be defined in such a way as to be expressible in both domains. In terms of path planning, routines to search route networks will involve examining graph concepts such as the Chinese Postman Problem to produce a path that will traverse each road within the network, the Travelling Salesman Probplem to produce paths that will overfly all raod junctions within the network, together with techniques to survey open areas with multiple UAVs. Decision making techniques using Temporal Logic to define the system behaviours will be presented
The resulting decision making and path planning techniques are illustrated by use of case studies that cover tracking contaminant clouds, monitoring highway traffic to surveying rural areas using UAVs and UGVs will be presented
In order to produce an autonomous system that has specific guaranteed behaviours, the link between the decision making code and the path planning needs to be defined in such a way as to be expressible in both domains. In terms of path planning, routines to search route networks will involve examining graph concepts such as the Chinese Postman Problem to produce a path that will traverse each road within the network, the Travelling Salesman Probplem to produce paths that will overfly all raod junctions within the network, together with techniques to survey open areas with multiple UAVs. Decision making techniques using Temporal Logic to define the system behaviours will be presented
The resulting decision making and path planning techniques are illustrated by use of case studies that cover tracking contaminant clouds, monitoring highway traffic to surveying rural areas using UAVs and UGVs will be presented