On the improvement of opposition-based differential evolution

Opposition-based Learning (OBL) is a new concept in machine intelligence, and has been proven to be an effective method to Differential Evolution (DE), Particle Swarm Optimization (PSO) and other population-based algorithms in solving many optimization problems. Opposition-based DE (ODE) is one of successful applications of OBL, which shows faster and more robust convergence than classical DE. In this paper, we focus on the improvement of ODE to enhance its performance on global optimization. The proposed approach, namely IODE, uses an improved OBL based on recombining current search point and another random point. The simulation results on 21 well-known benchmark functions show that the IODE outperforms ODE on the majority of test problems.