Banhan Bootsingha. Analysis of navigation of robot populations using finite automata models . Master's Degree(Applied Mathematics). King Mongkut's Institute of Technology North Bangkok. Central Library. : สถาบันเทคโนโลยีพระจอมเกล้าพระนครเหนือ, 2006.
Analysis of navigation of robot populations using finite automata models
Abstract:
Mobile rescue robots have been suggested as an alternative to human rescuers for
rescuing people in dangerous environments, e.g., people trapped by fire in a building.
In this thesis, the movements of a heterogeneous population of rescue robots in a
discrete environment are modeled using matrix methods. Two types of rescue
problem are considered. In the first type, the robots must travel from a specified finite
set of start states (rooms in a building) to a specified finite set of target rooms. If
possible, each target room should be reached by one robot. In the second type of the
problem, the robots must reach the target rooms and then return to a specified set of
safe rooms. The paths of each robot are classified using the concepts of blocked,
blocked but controllable and unblocked paths. The allowed motions of each robot are
modeled using automata theory. The transition function and transition diagram are
obtained. The transition diagram is then modeled to generate digraph and an
adjacency matrix for the digraph is constructed. An algorithm based on powers of the
adjacency matrix is presented which can derive blocked and controllability properties
for the paths of each robot. These properties of the population of mobile rescue
robots are then easily obtained for the two types of rescue problems. A Matlab
program based on the algorithm has been developed and tested on randomly selected
problems with a range of 3 to 30 robots and target rooms in buildings with a range of
10 to 300 rooms.