Analysis and prediction of lift failure parttern

With the construction of a large number of high-rise buildings in recent decades, the lift becomes one of the most frequently used communal facilities in people’s life. One tall building usually needs to be equipped with multiple lifts and the usage of lifts is also high. So, the huge number of tota...

Full description

Saved in:
Bibliographic Details
Main Author: Liu, Borui
Other Authors: Ling Keck Voon
Format: Thesis-Master by Coursework
Language:English
Published: Nanyang Technological University 2023
Subjects:
Online Access:https://hdl.handle.net/10356/172838
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:With the construction of a large number of high-rise buildings in recent decades, the lift becomes one of the most frequently used communal facilities in people’s life. One tall building usually needs to be equipped with multiple lifts and the usage of lifts is also high. So, the huge number of total lifts in a city as well as the frequent and heavy usage of lifts always comes with a great number of lift failures. As the maintenance company needs to deal with the problem of a large number of lift failures across different periods and locations every day, the efficient deployment of the maintenance resources in advance is particularly important. In this dissertation, the statistical technique is used to predict the health and operating conditions of the lift. By using techniques such as statistical data processing and machine learning to analyse the lift failure data which were collected over the past years, and then using algorithms to predict the number of lift failures in future months. It is envisaged that such techniques will be able to help the lift maintenance company to determine the specific deployment of the maintenance team to effectively reduce the average downtime of the lift when the lift failure happens. In addition, there are 4 presented simulation examples to demonstrate how the proposed method could predict future potential lift faults, as well as suggestions to improve the accuracy of our prediction model.