Optimization of RFID network planning for monitoring railway mechanical defects based on gradient-based Cuckoo search algorithm
Radio Frequency Identification (RFID) is an increasingly widespread and applied technology of automatic real-time monitoring and control railway assets. For that, the present research has developed an RFID network-planning model that can improve real-time information detection based on the tem...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English English |
| Published: |
2020
|
| Subjects: | |
| Online Access: | http://eprints.uthm.edu.my/4137/1/24p%20NIHAD%20HASAN%20TALIB.pdf http://eprints.uthm.edu.my/4137/2/NIHAD%20HASAN%20TALIB%20COPYRIGHT%20DECLARATION.pdf http://eprints.uthm.edu.my/4137/3/NIHAD%20HASAN%20TALIB%20WATERMARK.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Radio Frequency Identification (RFID) is an increasingly widespread and applied
technology of automatic real-time monitoring and control railway assets. For that, the
present research has developed an RFID network-planning model that can improve
real-time information detection based on the temperature and vibration in the gear and
motor of the train bogies. The selected system was Kuala Lumpur railway system,
which has been operating in the city of 243 km2 area. It involves three challenges
which represent the objectives of this thesis; the first is how to deal with the large�scale area and huge number of stations based on functional features. The second is
how to decide which station (or stations) is suitable to be applied with the RFID system
to help in monitoring the trains effectively. Finally, the third challenge is how to find
the optimal evolutionary method for railway network planning to increase the RFID
system performance. The solution strategy started in its initial input and process to find
effective stations that can serve the railway monitoring system well. The researcher
developed a new clustering model to separate the necessary data from unnecessary
data, and specified the suitable primary stations. For the second objective, the Analytic
Hierarchy Process (AHP) was used to decide which stations can be used to monitor
the railway system optimally. The Gradient-Based Cuckoo Search (GBCS) algorithm
was used to achieve the final objective. It solved the multi-objective functions of RNP
challenge. In the validation process, the results showed a superior finding compared to
the firefly algorithm. It was able to detect more tags by 3%, and a reduced number of
readers by 16.6%. In the large-scale area application, the GBCS algorithm achieved
100%, 93.75%, and 98.9% coverage for Maluri, Subang, and TBS stations,
respectively. In conclusion, this study presented a novel hybrid evolutionary algorithm
based on the combination of AHP with GBCS to specify optimal RFID reader
positions and amount based on the working train station domain. The present method
has proven its precise performance in RNP of large-scale area based on real-time
railway monitoring tasks. |
|---|