In recent decade, several nuture-inspired swarm intelligence-bused optimization techniques have been improved. These techniques, which give solutions close to optimum in un acceptable time, have been applied successfully to solve the problems in science and social sciences. Fireworks Algorithm (FA), inspired by observing fireworks explosion, is u new swarm intelligence algorithm. This relatively new technique has been utilized to tackle diverse problems and obtained better performance than other popular techniques.such us particle swarm optimization, ant colony, and genetic algorithm. Despite the good results obtained, FA requires Long computation time to achieve the optimum solution. To eliminate long computation time drawback of FA, in this study, a Fel using tournament selection IS proposed, The performance of the proposed IA, which involves tournament selection operator, is tested on well-known 15 numerical optimization problems. Experimentuf results reveal that proposed FA has u.s'ignificunt performance improvement in term of computation time and solution quality in comparison with original FA.