A C[formula omitted]-dominance-based solution estimation evolutionary algorithm for many-objective optimization

Abstract

Balancing convergence and diversity is a key issue for many-objective optimization problems (MaOPs), which is a great challenge to the classical Pareto-based multi-objective algorithms due to its severe lack of selection pressure. To relieve the above challenge, a Cα-dominance-based solution estimation evolutionary algorithm is proposed for MaOPs. In the proposed algorithm, a new dominance method, called Cα-dominance, is proposed to provide reasonable selection pressure for MaOPs. By designing a nonlinear function to transform the original objectives, Cα-dominance expands the dominated area where dominance resistant solutions located, while remains the solutions to be non-dominated in area close to Pareto optimal solutions. Furthermore, an adaptive parameter adjustment mechanism on the unique parameter α of Cα-dominance is designed to control the expansion degree of the dominance area based on the number of objectives and the stages of evolution. Finally, a new solution estimation scheme based on Cα-dominance is designed to evaluate the quality of each solution, which incorporates convergence information and diversity information of each solution. The experimental results on widely used benchmark problems having 5–20 objectives have shown the proposed algorithm is more effective in terms of both convergence enhancement and diversity maintenance.