Design method for low sensitivity optical systems based on a feedback optimization strategy

To minimize the overall complexity of assembly and alignment for optical systems during the design phase, it is essential to effectively control the internal tolerance allocation. However, conventional allocation methods rely heavily on empirical experience, rendering the alignment process time-consuming and lacking in theoretical guidance. To address this issue, this paper proposes a design method for optical systems with low error sensitivity, based on a feedback optimization strategy. The proposed method is applied to the design of an off-axis three-mirror anastigmat (TMA) system. With joint consideration of image quality and error sensitivity, the initial optical structure is generated by the non-dominated sorting genetic algorithm (NSGA-II). Furthermore, a feedback optimization mechanism is introduced to actively regulate the internal sensitivity distribution of the system. This strategy ensures effective control of error sensitivity, yielding an optical system with significantly reduced overall sensitivity. In simulation verification, the assembly-induced RMS wavefront error of the proposed system is reduced to 25% and 45% of that in the non-desensitized system and the system without feedback optimization, respectively. The results demonstrate that the proposed method rationally allocates internal sensitivity and significantly reduces overall sensitivity to assembly errors. Consequently, it effectively enhances the manufacturability and reliability of the optical system while maintaining excellent imaging performance.