A novel forward search strategy to automatically harmonize intensification and diversification in tabu search

摘要

Intensification strategy and diversification strategy are two highly important components of tabu search, which conflict each other in many cases. We proposed a novel forward search strategy that can encourage searching much thoroughly the neighborhood of elite solutions recorded, and can retain reasonable balance between intensification search and diversification search at the same time in order to get global optimal. Taking the classical combinatorial optimization problem TSP as samples, the conflict between intensification and diversification can be harmonized effectively by updating neighborhood, candidate and adjust tabu length automatically. Finally we give the experimental results.