A New Reinforcement Learning Algorithm with Fixed Exploration for Semi-Markov Control in Preventive Maintenance

Angelo Encapera; Abhijit Gosavi

doi:10.1115/MSEC2017-2880

A New Reinforcement Learning Algorithm with Fixed Exploration for Semi-Markov Control in Preventive Maintenance

Angelo Encapera
, Abhijit Gosavi

Engineering Management and Systems Engineering

Research output: Contribution to journal › Article › peer-review

Abstract

Artificial intelligence techniques can play a significant role in solving problems encountered in the domain of Total Productive Maintenance (TPM). This paper considers a new reinforcement learning algorithm called iSMART, which can solve semi-Markov decision processes underlying control problems related to TPM. The algorithm uses a constant exploration rate, unlike its precursor R-SMART, which required exploration decay. Numerical experiments conducted here show encouraging behavior with the new algorithm.

Original language	American English
Journal	Proceedings of the ASME 2017 12th International Manufacturing Science and Engineering Conference (2017, Los Angeles, CA)
Volume	3
DOIs	https://doi.org/10.1115/MSEC2017-2880
State	Published - Jun 1 2017

Keywords

Artificial intelligence techniques
Control problems
Learning algorithms
Manufacture
Markov processes
Numerical experiments
Preventive maintenance
Problem solving
Reinforcement learning
Semi-Markov controls
Semi-Markov decision process
Total productive maintenance

Disciplines

Operations Research, Systems Engineering and Industrial Engineering

Access to Document

10.1115/MSEC2017-2880License: Unspecified

https://doi.org/10.1115/MSEC2017-2880

Cite this

@article{b9ed4d2facf148038b3795bcf9c49bce,

title = "A New Reinforcement Learning Algorithm with Fixed Exploration for Semi-Markov Control in Preventive Maintenance",

abstract = " Artificial intelligence techniques can play a significant role in solving problems encountered in the domain of Total Productive Maintenance (TPM). This paper considers a new reinforcement learning algorithm called iSMART, which can solve semi-Markov decision processes underlying control problems related to TPM. The algorithm uses a constant exploration rate, unlike its precursor R-SMART, which required exploration decay. Numerical experiments conducted here show encouraging behavior with the new algorithm.",

keywords = "Artificial intelligence techniques, Control problems, Learning algorithms, Manufacture, Markov processes, Numerical experiments, Preventive maintenance, Problem solving, Reinforcement learning, Semi-Markov controls, Semi-Markov decision process, Total productive maintenance",

author = "Angelo Encapera and Abhijit Gosavi",

year = "2017",

month = jun,

day = "1",

doi = "10.1115/MSEC2017-2880",

language = "American English",

volume = "3",

journal = "Proceedings of the ASME 2017 12th International Manufacturing Science and Engineering Conference (2017, Los Angeles, CA)",

}

TY - JOUR

T1 - A New Reinforcement Learning Algorithm with Fixed Exploration for Semi-Markov Control in Preventive Maintenance

AU - Encapera, Angelo

AU - Gosavi, Abhijit

PY - 2017/6/1

Y1 - 2017/6/1

N2 - Artificial intelligence techniques can play a significant role in solving problems encountered in the domain of Total Productive Maintenance (TPM). This paper considers a new reinforcement learning algorithm called iSMART, which can solve semi-Markov decision processes underlying control problems related to TPM. The algorithm uses a constant exploration rate, unlike its precursor R-SMART, which required exploration decay. Numerical experiments conducted here show encouraging behavior with the new algorithm.

AB - Artificial intelligence techniques can play a significant role in solving problems encountered in the domain of Total Productive Maintenance (TPM). This paper considers a new reinforcement learning algorithm called iSMART, which can solve semi-Markov decision processes underlying control problems related to TPM. The algorithm uses a constant exploration rate, unlike its precursor R-SMART, which required exploration decay. Numerical experiments conducted here show encouraging behavior with the new algorithm.

KW - Artificial intelligence techniques

KW - Control problems

KW - Learning algorithms

KW - Manufacture

KW - Markov processes

KW - Numerical experiments

KW - Preventive maintenance

KW - Problem solving

KW - Reinforcement learning

KW - Semi-Markov controls

KW - Semi-Markov decision process

KW - Total productive maintenance

UR - https://scholarsmine.mst.edu/engman_syseng_facwork/633

UR - https://doi.org/10.1115/MSEC2017-2880

U2 - 10.1115/MSEC2017-2880

DO - 10.1115/MSEC2017-2880

M3 - Article

VL - 3

JO - Proceedings of the ASME 2017 12th International Manufacturing Science and Engineering Conference (2017, Los Angeles, CA)

JF - Proceedings of the ASME 2017 12th International Manufacturing Science and Engineering Conference (2017, Los Angeles, CA)

ER -

A New Reinforcement Learning Algorithm with Fixed Exploration for Semi-Markov Control in Preventive Maintenance

Abstract

Keywords

Disciplines

Access to Document

Other files and links

Fingerprint

Cite this