An Optimization Method for Dual-arm Robot Task Scheduling Considering Collision Risk

Zixuan Chen,Xianzhe Yuan,Qiang Gu,Chunxu Hu,Dingxin He
DOI: https://doi.org/10.1109/cac57257.2022.10054947
2022-01-01
Abstract:An optimization method is presented for dual-arm robot task scheduling in this paper. The purpose of task scheduling is to assign tasks and determine optimal task paths for dual-arm robot. The robot tasks are considered as a set of task points. The task scheduling is modeled as a Multi-Objective Multiple Traveling Salesman Problem (MOMTSP) to optimize three goals (total time consumption, takt time and collision risk). An improved algorithm based on Non-dominated Sorting Genetic Algorithm (NSGA-II) is proposed for the MOMTSP solution. The simulation demonstrates that the improved algorithm performs faster iterations and better results. Experimental results are shown to prove the method improvement compared with the classical method based on genetic algorithm. The proposed method significantly reduces the collision risk and time consumption.
What problem does this paper attempt to address?