A Cross Strategy-Based Agent-Assisted Multi-Population Particle Swarm Optimization Algorithm

With the rapid increase in the complexity of modern engineering systems and scientific computing problems, Large-Scale Expensive Optimization Problems (LSEOPs) have become a core challenge in the field of optimization. Traditional Evolutionary Algorithms (EAs) encounter the "curse of dimensionality" in high-dimensional scenarios, and most of the existing Surrogate-Assisted Evolutionary Algorithms (SAEAs) are designed for low-dimensional problems and struggle to adapt to the characteristics of high-dimensional data. In this paper, a Cross Strategy-Based Agent-Assisted Multi-Population Particle Swarm Optimization algorithm (CS-AAMPSO) is proposed to improve the efficiency and robustness of methods for working out LSEOPs, and the experimental results show that the performance is improved to some extent.