History-based response threshold model for division of labor in multi-agent systems

Wonki Lee, Daeeun Kim

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Dynamic task allocation is a necessity in a group of robots. Each member should decide its own task such that it is most commensurate with its current state in the overall system. In this work, the response threshold model is applied to a dynamic foraging task. Each robot employs a task switching function based on the local task demand obtained from the surrounding environment, and no communication occurs between the robots. Each individual member has a constant-sized task demand history that reflects the global demand. In addition, it has response threshold values for all of the tasks and manages the task switching process depending on the stimuli of the task demands. The robot then determines the task to be executed to regulate the overall division of labor. This task selection induces a specialized tendency for performing a specific task and regulates the division of labor. In particular, maintaining a history of the task demands is very effective for the dynamic foraging task. Various experiments are performed using a simulation with multiple robots, and the results show that the proposed algorithm is more effective as compared to the conventional model.

Original languageEnglish
Article number1232
JournalSensors (Switzerland)
Volume17
Issue number6
DOIs
Publication statusPublished - 2017 Jun

All Science Journal Classification (ASJC) codes

  • Analytical Chemistry
  • Biochemistry
  • Atomic and Molecular Physics, and Optics
  • Instrumentation
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'History-based response threshold model for division of labor in multi-agent systems'. Together they form a unique fingerprint.

  • Cite this