OPTIMIZATION ALGORITHMS FOR NODE ALLOCATION IN VISION GUIDED MULTI-ROBOT PATTERN FORMATION

Ranganathan Senthilnathan, Ganesan Vignesh, Arjun Venugopal, Selvakumaran Anish Kanna, and Sundaramoorthy Sidhanathan

Keywords

Pattern formation, node allocation, multi-robot systems, vision guidance, optimization, genetic algorithm

Abstract

In multi-robot systems, the instantaneous location of each robot is generally assigned based on some pre-determined criteria, which help the collective mission of the robots. Assigning the destination nodes may be considered an optimization problem, as the process must be based on either maximizing or minimizing a cost function aiding in the mission accomplishment. The paper presents the details of two optimization algorithms for the node allocation problem in an overhead-camera based vision guided multi-robot system intended for geometrical pattern formation. The first method is tailor-made non-heuristic based, and the second method is based on genetic algorithm optimization. The optimization algorithms address the problem of assigning destination nodes to multiple mobile robots such that their location in the respective nodes assigned would ensure proper formation of the intended pattern. The cost function is based on the total distance travelled by the robots. The details of the time performance and optimality of the total distance travelled by the robots are presented in this paper. Based on the nodes allocated by the optimization algorithms, iterative global trajectory planning is executed for every iteration, where each iteration is based on the recent update of the image of the scene.

Important Links:



Go Back