Empirical Analysis of Improvement Algorithms for the Pickup and Delivery Traveling Salesman Problem

Abstract

This paper presents an empirical analysis of some implementations of tour improvement heuristics designed for the pickup and delivery traveling salesman problem. Results show that while the classical 2-opt traveling salesman algorithm can be implemented very efficiently, available implementations of the 3-opt are much less efficient. Among tested algorithms, the only one to produce significant improvements on large scale problems is our adaptation of the 4-opt improvement heuristic.

3 Figures and Tables

Cite this paper

@inproceedings{Renaud2004EmpiricalAO, title={Empirical Analysis of Improvement Algorithms for the Pickup and Delivery Traveling Salesman Problem}, author={Jacques Renaud and Fayez F. Boctor}, year={2004} }