Chaos-enhanced self-adaptive particle swarm optimization with simulated annealing for digital lithography mask optimization

In this paper, an efficient approach to mask optimization for digital micromirror device lithography is proposed, leveraging an enhanced particle swarm optimization algorithm, which significantly elevates the resolution and precision of lithography. Initially, chaos mapping is applied to the initial population to enhance particle diversity, thereby improving the optimization efficiency of the algorithm. Subsequently, self-adaptive parameter adjustments and simulated annealing are integrated to effectively avoid premature convergence and escape local optima. Numerical simulation results demonstrate a substantial reduction in pattern errors between the printed and the target images by 95.2%, 95.4%, and 89.2%. The proposed algorithm markedly surpasses conventional optimization methods, notably bolstered in optimization efficiency and pattern accuracy.