Premium
Adaptive Fuzzy Collaborative Task Assignment for Heterogeneous Multirobot Systems
Author(s) -
Zhang Lu,
Zhong Hao,
Nof Shimon Y.
Publication year - 2015
Publication title -
international journal of intelligent systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.291
H-Index - 87
eISSN - 1098-111X
pISSN - 0884-8173
DOI - 10.1002/int.21725
Subject(s) - robot , computer science , task (project management) , variety (cybernetics) , fuzzy logic , set (abstract data type) , artificial intelligence , distributed computing , machine learning , engineering , systems engineering , programming language
Heterogeneity in robot model mix is advantageous in emerging and future robotic applications. For instance, in a fully automated industrial package loading and unloading scenario, a variety uncertain types of packages need to be continuously sorted and loaded onto designated trucks. To handle the variation of tasks, multiple types of robots, or heterogeneous robots, are designed in the system. However, in such a collaborative system consisting of heterogeneous robots, ineffective task assignments often lead to bad collaboration and thus poor efficiency. To improve the robot collaboration, we herein define the collaborative task assignment problem and develop a fuzzy collaborative intelligence based algorithm to optimize the assignment plans. Specifically, the collaboration type, the collaboration matrix, and the assignment matrix are defined and the new algorithm for adaptive fuzzy collaborative task assignment relies on intuitionistic fuzzy set theory. An unsorted package loading and unloading tasks by heterogeneous robots is used as a case example to validate the new algorithm. Experiments indicate with statistical significance that the new approach shortens total completion time by 21%, reduces total energy consumption by 23%, and increases loading accuracy by 31%, compared with the traditional static task assignment method commonly practiced. The developed approach can be applied to different emerging collaborative systems to improve systems’ collaborative intelligence.