Adaptive genetic algorithm for path planning of loosely coordinated multi-robot manipulators

Abstract

Adaptive genetic algorithm ASAGA, a novel algorithm, which can dynamically modify the parameters of Genetic Algorithms in terms of simulated annealing mechanism, is proposed for path planning of loosely coordinated multi-robot manipulators. Over the task space of a multi-robot, a strategy of decoupled planning is also applied to the evolutionary process, which enables a multi-robot to avoid falling into deadlock and calculating of composite C-space. Finally, two representative tests are given to validate ASA GA and the strategy of decoupled planning.