A reinforcement learning approach to optimal part flow management for gas turbine maintenance. (February 2020)

• Record Type: Journal Article
• Authors: Compare, Michele; Bellani, Luca; Cobelli, Enrico; Zio, Enrico; Annunziata, Francesco; Carlevaro, Fausto; Sepe, Marzia
• Is Part Of: Proceedings of the Institution of Mechanical Engineers. Volume 234:Number 1(2020:Feb.)
• Page Start: 52
• Page End: 62
• Publication Date: 2020-02
• Subjects: Part flow -- reinforcement learning -- gas turbine; Reliability (Engineering) -- Mathematical models -- Periodiclals; Risk assessment -- Mathematical models -- Periodicals; Engineering design -- Mathematical models -- Periodicals; 620.00452
• DOI: 10.1177/1748006X19869750
• Languages: English
• ISSNs: 1748-006X

Abstract

We consider the maintenance process of gas turbines used in the Oil and Gas industry: the capital parts are first removed from the gas turbines and replaced by parts of the same type taken from the warehouse; then, they are repaired at the workshop and returned to the warehouse for use in future maintenance events. Experience-based rules are used to manage the flow of the parts for a profitable gas turbine operation. In this article, we formalize the part flow management as a sequential decision problem and propose reinforcement learning for its solution. An application to a scaled-down case study derived from real industrial practice shows that reinforcement learning can find policies outperforming those based on experience-based rules.