A fast evolutionary algorithm for dynamic bi-objective optimization problems
Min Liu, Wenhua Zeng · 2012
Many real-world optimization problems involve multiple objectives, constraints, and parameters which constantly change with time. In this paper, we suggest a fast dynamic bi-objective evolutionary algorithm (DBOEA). Specifically, a fast bi-objective non-dominated sorting is introduced to reduce the cost of the layering of non-dominated fronts. A differential evolution operator is also adopted as the new evolutionary search engine so as to accelerate the optimization search speed and improve the obtained results. The DBOEA is very fit for dynamic bi-objective optimization, for its computational complexity is O(N log N). The simulate results demonstrate that the proposed DBOEA outperforms the well-known dynamic non-dominated sorting algorithm II (DNSGA-II) not only in running speed, but also in terms of finding a diverse set of solutions and in converging near the dynamic Pareto optimal front.