Multi-objective Optimization with Modified Pareto Differential Evolution

In this paper, a modified Pareto differential evolution algorithm approach is introduced to solve multi-objective optimization problems. The modified approach including: improving the crowding mechanism by introducing the Niche theory, introducing a time variant scaling factor F, and using a simple but efficient mutation operator which combines information of the both better and worse solution in the population. The performance of the MDE is measured on the set of ZDT benchmark functions on multi-objective optimization.