Please use this identifier to cite or link to this item:
Title: Efficiency analysis of tabu search heuristic for static dial-a-ride problem
Authors: Ho, Song Guang
Keywords: DRNTU::Engineering::Electrical and electronic engineering
Issue Date: 2018
Abstract: Multi-vehicle routing has become increasingly important due to the rapid development in autonomous vehicle technology and Internet of Things. By developing better multivehicle routing algorithms, the current transportation system can be made more efficient and travel time can be significantly reduced. A future-proof algorithm must be one that utilize all available data (i.e. traffic condition, weather condition) to make routing algorithm more accurate and efficient. There are many algorithms developed to solve diala-ride problem. One of them is tabu search algorithm developed for dial-a-ride problems by Cordeau and Laporte in 2003. This thesis investigates ways to reduce the computational time of tabu search algorithm by proposing an improved tabu search algorithm. The efficiency validation of the algorithm has been performed using twelve variants of the major parameters. These variants are based on the type of insertion techniques, neighborhood evaluation methods and time window adjustments used. There are two types of insertion techniques: one-step and two-step insertion; there are three types of neighborhood evaluation techniques: one-level, two-level and three-level neighborhood evaluation. The effect of implementing time window adjustment is also tested. From the extensive simulations, it has been identified that three-level neighborhood evaluation technique and time window adjustment are crucial to improve the efficiency. One-step and two-step insertion both have their own advantages, and should be applied according to application specifications and needs. Number of iteration for the tabu search algorithm should also be adjusted accordingly to reduce the computation time.
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Theses

Files in This Item:
File Description SizeFormat 
HO SONG GUANG_ 2017.pdf
  Restricted Access
7.72 MBAdobe PDFView/Open

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.