Optimal routing control of a construction machine by deep reinforcement learning

Zeyuan Sun, Masayuki Nakatani, Yutaka Uchimura

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Deep reinforcement learning algorithms are rapidly growing, and expected to be applied to many industrial fields. In this paper, we proposed a method that combines a deep Q-network with batch normalization to generate an optimal route for a grading machine. The goal is to achieve autonomous operation of the grading machine. For the learning platform, a grading simulator was developed to emulate the grading work. The proposed method was evaluated with the grading simulator, and showed better route searching performance results than the conventional method.

Original languageEnglish
Title of host publicationProceedings - 2018 IEEE 15th International Workshop on Advanced Motion Control, AMC 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages187-192
Number of pages6
ISBN (Electronic)9781538619469
DOIs
Publication statusPublished - 2018 Jun 1
Event15th IEEE International Workshop on Advanced Motion Control, AMC 2018 - Tokyo, Japan
Duration: 2018 Mar 92018 Mar 11

Publication series

NameProceedings - 2018 IEEE 15th International Workshop on Advanced Motion Control, AMC 2018

Other

Other15th IEEE International Workshop on Advanced Motion Control, AMC 2018
Country/TerritoryJapan
CityTokyo
Period18/3/918/3/11

Keywords

  • Artificial intelligence
  • Batch normalization
  • Deep reinforcement learning
  • Grading machine

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Mechanical Engineering
  • Control and Optimization

Fingerprint

Dive into the research topics of 'Optimal routing control of a construction machine by deep reinforcement learning'. Together they form a unique fingerprint.

Cite this