An efficient fitness-based differential evolution algorithm and a constraint handling technique for dynamic economic emission dispatch

In this paper, an efficient fitness-based differential evolution (EFDE) algorithm and a constraint handling technique for dynamic economic emission dispatch (DEED) are proposed. In EFDE, there are three improvements compared to the standard differential evolution (DE) algorithm. First, an archive containing the current and previous population is established to provide more candidate solutions. Second, two mutation strategies are used to generate mutant individuals, where the population similarity is introduced to choose a suitable one between DE/rand/1 and DE/best/1. The fitness-based mutation operation is efficient to balance the exploration and exploitation ability of EFDE. Third, EFDE adopts a random-based mutation factor, and the crossover rate with the learning ability is developed to produce more excellent solutions. In addition, the infeasible solutions can be effectively avoided by the proposed repair technique. Four cases are selected to judge the performance of the proposed EFDE and constraint handling technique. For the fuel cost and emission minimizations of four DEED cases, a normalized approach (NA) is used to help EFDE to find the best compromise solutions in the evolution process. According to the simulation results, EFDE exhibits a huge advantage in comparison with the other approaches for the single-objective and multi-objective optimization problems.