Adaptive hybrid response prediction for dynamic multi-objective optimization

Dynamic multi-objective optimization involves a wide range of problem types, presenting substantial challenges to the adaptability of optimization algorithms as the complexity of problems grows: particularly when the environment undergoes irregular and drastic changes, most algorithms fail to account for appropriate responses to the degree of environmental change, making it difficult to better solve such dynamic multi-objective optimization problems. To address these challenges, this paper proposes an adaptive hybrid response prediction algorithm (AHRPA). Specifically, the change response strategy includes partitioning prediction strategy and enhanced adaptive randomization strategy. Under different degrees of environmental changes, the population proportion generated by different response mechanisms is dynamically adjusted, thereby achieving rapid convergence of the population and enhancement of diversity. The algorithm is systematically compared with several state-of-the-art algorithms across a variety of benchmark problems. The empirical analysis demonstrates that the proposed AHRPA algorithm exhibits highly competitive performance on the majority of benchmark problems.