A Genetic Algorithm for tasks allocation and sequencing in a human robot assembly system

Abstract

Due to the increased productivity and flexibility requirements of production lines, research and industry are increasingly interested in the integration of collaborative human-robot systems (CHR) within flexible assembly lines; which will allow combining human and robotic capabilities. In this work, we focused on the design of a control system that enables the planning / coordination of human-robot collaboration (CHR) tasks. The system is based on an optimization process that resolves the affectation and the sequencing problem (permutation of tasks) between the robot and operator. Considering that Tree-based assembly tasks constraints must to be taken into account during the resolution, a novel encoding-based genetic algorithm is proposed for the resolution of the problem. In the proposed method, the genetic algorithm is implemented using an indirect (non-permutation) encoding scheme and a dedicated evaluation mechanism that avoid from any kind of solution repair after recombination operators. Preliminary results validate the approach on generated benchmarking instances