A modified Marine predators algorithm for solving single- and multi-objective combined economic emission dispatch problems

Decreasing the emission of greenhouse gases and the fuel cost are the main principles of the combined economic emission dispatch (CEED) optimization problem. In this work, a modified version of the Marine Predators Algorithm (MMPA), is proposed and applied for solving single- and bi-objective CEED problems. MMPA is suggested to improve the performance of conventional MPA. It includes a comprehensive learning approach to share the best experiences among all the individuals to avoid premature convergence. The MMPA increases the populations' effectiveness to reach the optimum fitness solution and this is helpful in cases of single- and bi-objective functions. The proposed MMPA has been mathematically assessed with twenty-eight 50-dimensional benchmarks of CEC2017. Also, it is successfully tested on four electrical power generation systems (each of them with 3, 5, 6, and 26 generating units respectively). The results achieved by MMPA are compared with those obtained by the original MPA and other recent optimization methods. Concerning the bi-objective problem, the Pareto approach is combined with the proposed MMPA and conventional MPA techniques to reach a group of non-dominated (ND) solutions, and then the fuzzy method is utilized to choose the best compromise solution. The reached results display that the proposed MMPA technique gives better performance than the conventional MPA and the others.