A new evolutionary algorithm for global numerical optimization

In this paper, a new evolutionary algorithm referred to as importance search algorithm (ISA) is designed to solve global numerical optimization problems with continuous variables. The proposed algorithm mainly consists of initialization process and iteration process in which the initialization process is used to initialize a population of random feasible solutions, and the iteration process is accomplished according to the move of the best particle in the colony. To show the effectiveness of the proposed ISA for global numerical optimization problems, it is applied to solve some typical benchmark test functions which are widely used in the literature and compared with the computational results obtained by using particle swarm optimization (PSO). The comparative results show that the proposed ISA is more effective than PSO when the problems with a large dimensions and it can And the optimal or near-optimal solutions of global numerical optimization problems.