The objective of this study is to develop a tracking algorithm which would allow a biologist to locate tagged animals using an unmanned aerial vehicle. The algorithm is developed to track the red wolves in a wildlife refuge in North Carolina. The red wolf is an endangered species and there are about a hundred of them living in the wild. For tracking purposes, each animal is outfitted with a collar containing a radio transmitter which emits a signal of a specific frequency over a given range. Using transmitter signal range, we discretize the refuge terrain map and develop a suboptimal greedy algorithm to search for the probabilistic targets striving to minimize the expected traveling time. To reduce the computational complexity, we consider a two-layered approach to the search problem. At the top level, the algorithm chooses the pack sectors as targets and, at the lower level, employs a local policy to travel between the discrete sensing locations. Using the animal behavior model provided by the wildlife biologists, we develop the probability rules to update the predicted target locations given the location of already found wolves. Finally, we present the results of a numerical simulation.
|Original language||English (US)|
|Title of host publication||Proceedings of the International Conference on Automation Technology for Off-road Equipment, ATOE 2004|
|Number of pages||9|
|State||Published - Dec 1 2004|