实数编码自适应遗传算法及其在光瞳滤波器参数控制中的应用

Real-coded adaptive genetic algorithms and its application in parameter control of pupil filter

提出一种基于实数编码自适应遗传算法的光瞳滤波器光瞳函数参数优化设计的新方法.通过计算机模拟,结合实际使用场合对光能效率、旁瓣影响因子和超分辨能力的要求,选择合适参数实现最佳轴向超分辨成像效果.利用遗传算法设计的光瞳滤波器设计简单,有较好的综合效果,不仅具有较小的u0和M值,而且有较大的S值.理论和实践均表明该方法是高效和可靠的.

A new optimization approach to function parameters of pupil filter based on adaptive real-coded genetic algorithm is proposed. Out of consideration of practical need for optical energy efficiency, sidelobe intensity, and superresolution, the appropriate parameters are selected by means of the computer simulation to achieve the best axial superresolution. By using adaptive real-code genetic algorithms, the design of pupil filters becomes simpler and possesses better comprehensive effects. The pupil filters designed with the aid of adaptive real-code genetic algorithm possess smaller u_0,M and larger S. It is verified by the theory and practice that this method is highly effective and reliable.