A novel evolutionary algorithm for nonlinear programming problems

This paper proposes a novel evolutionary algorithm referred to as importance search algorithm (ISA) for constrained nonlinear programming problems, which is initialized with a population of random feasible solutions and searches for the optimal solution by updating generations. The ISA mainly consists of initialization process and iteration process, and the process of iteration is accomplished according to the move of the best particle in the colony. To show the effectiveness of the proposed ISA, we apply it to solve 8 different kinds of nonlinear programming problems, and compare the computational results with those obtained by using particle swarm optimization (PSO) and genetic algorithm (GA) in the literature. The comparison results show that the ISA is efficient to the problems in multiple-dimensional, nonlinear and complex programming problems. Furthermore, three test problems are selected to demonstrate the effectiveness of the ISA from the sensitivity perspective. The numerical experiments show that the ISA is robust to the parameters settings.