Edge recombination with edge sensitivity in TSP problem

We use an evolutionary algorithm with a modified version of edge recombination crossover to search near optimal tours in the Traveling Salesman Problem (TSP). We introduce so called edge sensitivity to estimate the importance of having a given edge within an optimal solution. In order to test this approach we used the data files gr24, bays29 and gr48. Results we obtained show that edge sensitivities allow us to find solutions quicker than other similar algorithms with an edge recombination crossover.