Show simple item record

dc.contributor.authorGascon, Martin
dc.contributor.authorKumar, Nikhil
dc.contributor.authorGhosh, Rana
dc.date.accessioned2022-02-04T14:15:17Z
dc.date.available2022-02-04T14:15:17Z
dc.date.copyright2020/02/24/
dc.date.issued2020
dc.identifier.issn0195-0738
dc.identifier.otherjert_142_7_070908.pdf
dc.identifier.urihttp://yetl.yabesh.ir/yetl1/handle/yetl/4273282
dc.description.abstractThere are new challenges for plant operators due to the increased share of renewable energy. Plant operators must maintain high reliability and high profits while plants are being required to be more flexible to compensate for the variable generation addition of these renewables into the grid. Plant operators must deal with the thermal strain and the wear-and-tear of such operations. Various models have been proposed in the literature. However, no work has been reported on the development of a robust prediction model. The aim of this study was to determine which machine learning algorithm gives the best estimation of boiler component remaining useful life using plant operations. The flexible operation for all units was estimated using the Intertek hourly MW analysis and damage modeling software Loads Model™. We used several plant features as predictors (such as equipment manufacturer, operating regime, and ramp rates). We tested five different machine learning techniques and found that gradient boost is the best approach to predict the reduction in life span of the plant with over 90% precision.
dc.publisherThe American Society of Mechanical Engineers (ASME)
dc.titlePredicting Power Plant Equipment Life Using Machine Learning
dc.typeJournal Paper
dc.journal.volume142
dc.journal.issue7
dc.journal.titleJournal of Energy Resources Technology
dc.identifier.doi10.1115/1.4044939
dc.page70908
dc.treeJournal of Energy Resources Technology:;2020:;volume( 142 ):;issue: 007
dc.contenttypeFulltext


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record