Özer, Emre (2020) Predictive maintenance algorithm design for bus door systems. Design Project, Mechanical Engineering.
![]() |
Text
toestemming.pdf Restricted to Registered users only Download (122kB) |
|
|
Text
mMe_2020_OzerE.pdf Download (9MB) | Preview |
Abstract
The decrease of costs for condition monitoring equipment created a grand opportunity to inspect the state of a system for improved operations and increased efficiency. A major application for CM equipment is error detection and accordingly, error prevention. The enhanced computational power enabled the engineers to estimate the overall effective life span for machines leading to preventive maintenance. The emerging CM technologies promise a more powerful technique called predictive maintenance. Constantly monitoring the system provides information about the operations of the machine and enables to foresee any kind of failures. This study aims to prepare a base algorithm for a predictive maintenance system, planned to be implemented in bus doors. The data collected by Ventura Systems via an endurance test will be analysed to observe the correlations between parameters and to detect patterns. Thereafter based on the initial analysis, a new test set will be designed to gather more useful information. Finally, when the test is done and the data is collected, a predictive model will be generated. Three different prediction models will be created to evaluate the reliability of the design. The goal is to inspect the data from the endurance test, to build a predictive model with that data and to investigate the correlations of the data gathered from different sensors. Performance optimization and external physical effects are out of the scope of the project and will not be included.
Item Type: | Thesis (Design Project) |
---|---|
Supervisor name: | Cao, M. and Jayawardhana, B. |
Degree programme: | Mechanical Engineering |
Thesis type: | Design Project |
Language: | English |
Date Deposited: | 15 Jan 2021 08:55 |
Last Modified: | 30 Jul 2021 12:18 |
URI: | https://fse.studenttheses.ub.rug.nl/id/eprint/23692 |
Actions (login required)
![]() |
View Item |