Improving Travelling Salesman Problem Solution Diversity Using Graph Based Evolutionary Algorithms
Jayakanth Jayachandran, Steven Michael Corns · ASME Press eBooks · 2010
This project investigates the use of Graph-Based Evolutionary Algorithms in generating diverse solutions for the traveling salesman problem. The traveling salesman problems have a number of applications in science and engineering. Using Graph-Based Evolutionary Algorithms to generate a set of diverse solutions to this problem will provide an opportunity to explore how resiliency can be added to applications such as supply chain management problems. A dissimilarity measure based on transpositions is used to compute the diversity of the solutions. It was found that the number of diverse solutions generated by multiple algorithm runs increases with the rate of information shared within the algorithm.