This paper deals with anomaly detection from collected flight data for a Boeing 757-200currently in use by Icelandair. Previous work on anomaly detection using generated datahas been shown to give good results using multiple linear regression models. Using a similarapproach the paper addresses how the amount of data effects the models ability to detectanomalies. Simple regression approach is compared to a quadratic one and practical use ofmaintenance data in this context is explored. Finally different model variations are investigatedto attempt further root cause analysis of faults.
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:kth-239042 |
| Date | January 2018 |
| Creators | Sveinbjarnarson, Unnar Már |
| Publisher | KTH, Flygdynamik |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
| Relation | TRITA-SCI-GRU ; 2018:269 |
Page generated in 0.0017 seconds