A Scatter Search Variant to Solve max-SAT Problems

In this work, a new scatter search-based approach is studied for the NP-Hard satisfiability problems, in particular for its optimization version namely max-SAT. The paper describes a scatter search algorithm enhanced with a tabu search component combined with a uniform crossover operator. The latter is used to identify promising search regions while tabu search performs an intensified search of solutions around these regions. Our objective is to achieve a good compromise between intensification and diversification in the search process. Empirical tests are performed on DIMACS benchmark instances. Numerical results show a considerable performance in favor of our new approach. keywords: scatter search, tabu search, max-SAT, metaheurictics, hybridization.