University of Twente Student Theses

Login

Using Granger Causality to generate early warnings from Patrol Vessel platform sensor data

Wieringa, J. (2024) Using Granger Causality to generate early warnings from Patrol Vessel platform sensor data.

[img] PDF
6MB
Abstract:The Royal Netherlands Navy (RNLN), will undergo various replacement programs in the coming 20 years. To improve on availability of the systems on-board the need for more sophisticated maintenance systems has arisen. With the goal of creating smart maintenance (SM) on the ships, a method called Granger causality (GC) was proposed. The method, originating from econometrics, aims to evaluate the causality between two parameters. Datasets selected from archived sensor readings of the Oceangoing Patrol Vessels in the RNLN are used for the analysis. In selected use-cases, in- and output parameters were identified for the application of the method. Causality between in- and output parameters could then confirmed in a normal working condition of a propulsion system. With the assumption that GC is lost during malfunction a detection system was proposed and applied on a larger dataset. Applications showed that with a performance expressed in a F1-score ranging between 0.88 and 0.96, visually identified malfunctions could be detected. The thesis therefore proved the feasibility of the application of GC in a naval setting.
Item Type:Essay (Master)
Clients:
Ministerie van defensie, Utrecht, Netherlands
Faculty:ET: Engineering Technology
Subject:52 mechanical engineering
Programme:Mechanical Engineering MSc (60439)
Link to this item:https://purl.utwente.nl/essays/102558
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page