New collision avoidance scheme for multi-agents: a solution to the blindman's problem
Bibhya Nand Sharma, Jito Vanualailai, J-H. Ha, Avinesh Prasad · 2009
The Lyapunov-based control scheme is applied to construct new continuous control laws for multi agents to perform a number of subtasks integral to the motion planning and control problem, in a priori known environment. The subtasks include navigation in the dynamic environment; respecting kinodynamic constraints; parking; and being aligned to a given final orientation. In this paper, we will incorporate a new subtask classified as the blindman’s problem that will require avoidance of a special category of dynamic obstacles. This category involves dynamic obstacles that are contained within a framework of unique constraints. A requirement of this interesting problem is that the avoidance maneuvers be executed by the mobile robots only and not by the dynamic obstacles. To the authors’ knowledge, this is the first ever attempt to design a one-way avoidance scheme to solve the blindman’s problem. In this work, the minimum distance technique (MDT) coupled with tuning parameters will be instrumental in forcing the desired orientations in the final configuration. Computer simulations are used to illustrate the effectiveness of the proposed controllers.