Controlling an autonomous agent for exploring unknown environments using switching prelearned modules

Takahito Hata, Masanori Suganuma, Tomoharu Nagao

Research output: Contribution to journalArticlepeer-review

Abstract

In this paper, we try to acquire various behavior patterns of autonomous exploration agent using several learning environments. In case of previous learning methods using a single behavior rule set, it is hard to acquire the behavior that covers all learning environments. In our method, we divide learning environments into some primitive environments whose properties differ each other, and then generate modules that are specialized for each primitive environment. To optimize behavior rules of agents, we adopt Graph Structured Program Evolution (GRAPE) which can automatically generates graph structured programs. In unknown environments, each module is switched by a program named "switcher". The switcher selects the module that acts better in a neighboring environment. Through several experiments, our method achieved higher exploration rate in unknown environments compared to simple GRAPE, random search, and the method that switches modules randomly.

Original languageEnglish
Pages (from-to)157-164
Number of pages8
JournalIEEJ Transactions on Electronics, Information and Systems
Volume138
Issue number2
DOIs
Publication statusPublished - 2018 Jan 1
Externally publishedYes

Keywords

  • Automatic programming
  • Autonomous agent
  • Generalization
  • Genetic programming
  • Modulalization

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Controlling an autonomous agent for exploring unknown environments using switching prelearned modules'. Together they form a unique fingerprint.

Cite this