Evolutionary algorithms have gradually been recognized as effective approaches for solving many-objective optimization problems, owing to their population-based search mechanisms and inherent parallelism. To address the challenge of simultaneously achieving convergence and diversity in many-objective optimization, a many-objective evolutionary algorithm assisted by dual-indicator evaluation, termed MaOEA-ASDE, is proposed. First, in the mating selection stage, a dynamic mating selection strategy assisted by the \(I_{\varepsilon + }\) indicator and the ISDE+ indicator is designed. Through this strategy, high-quality parent individuals are adaptively and preferentially selected. Second, in the environmental selection stage, a two-step cooperative mechanism is developed to retain offspring with good convergence and diversity. This mechanism includes a screening strategy guided by the \(I_{\varepsilon + }\) indicator and a regulation and pruning strategy assisted by the ISDE+ indicator. In the first step, candidate solution pairs with similar search performance are identified from the nondominated solution set under the guidance of the \(I_{\varepsilon + }\) indicator. This process helps ensure the exploration capability of the population across the entire objective space. In the second step, these candidate pairs are further refined according to the ISDE+ indicator, and individuals with smaller ISDE+ values are eliminated, thereby enhancing convergence. Finally, extensive comparative simulation experiments are conducted on the DTLZ and MaF benchmark test suites, which involve a large number of objectives and include both regular and irregular Pareto fronts. The proposed algorithm is compared with six state-of-the-art many-objective evolutionary algorithms. The experimental results demonstrate that the proposed algorithm exhibits strong robustness and adaptability in balancing convergence and diversity for many-objective optimization problems.
