An Effective Algorithm based on Search Economics for Multi-Objective Optimization

Abstract

An effective multi-objective search algorithm based on a new meta-heuristic algorithm, named search economic (SE), is presented in this study. The basic idea of SE is to first partition the solution space into a certain number of regions to keep the diversity. Then, it will determine the later search directions by the so-called expected value that is composed of the objective values of the best-so-far solution of each region, the searched solutions, and the number of searches invested on a region. More important, the proposed algorithm will invest limited computing resources on promising regions to find a better Pareto optimal set (POS). Different from other search economics-based algorithms, the proposed method uses two transition operators of differential evolution and adds a self adaptive mechanism to tune its parameters. Experimental results show that the proposed algorithm outperforms all the other metaheuristic algorithms compared in this study in most cases in the sense that it can get a more uniformly distributed POS and a smaller distance to the Pareto optimal front.