Particle Swarm Optimization Algorithm for Visible Light Transmission Performance

Abstract

With the development of precision imaging, micro-nano optics, and intelligent manufacturing, complex optical systems are placing higher demands on image quality, cost, and manufacturability. Traditional design processes based on damped least squares and algorithms such as genetics and differential evolution are becoming increasingly inadequate in terms of global optimization and automation. This paper focuses on the application of particle swarm optimization (PSO) in optical model optimization. It uses a combination of literature review and comparative analysis to summarize the research progress of PSO in lens design, freeform and aspherical surface optimization, optical thin films, and structural colors, as well as illumination and antenna optical systems in the past five years. Starting with standard PSO, adaptive PSO, multi-objective PSO, and hybrid frameworks such as PSO + local optimization and PSO + intelligent algorithm. This paper summarizes their advantages in search efficiency, solution diversity, and design automation, and analyzes problems such as premature convergence, high computational cost, parameter sensitivity, and insufficient embedding of physical constraints. Finally, it proposes development directions such as physical information-driven approaches, surrogate model acceleration, and multi-algorithm hybrid optimization, providing a reference for the research and engineering application of optical intelligent design methods.