变间距采样及准梯度下降法实现光束整形  

Shaping by Using Quasi Gradient Decent Algorithm at Variable Sampling Intervals

作　　者：赵逸琼[1] 伍源 周秋桂[1] 李永平[1] 

机构地区：[1]中国科学技术大学物理系,安徽合肥230026 [2]国家同步辐射实验室,安徽合肥230027

出　　处：《计算物理》2004年第1期10-14,共5页Chinese Journal of Computational Physics

基　　金：国家863-804课题资助项目

摘　　要：　利用输入输出不同的变间距采样通过汉克尔变换实现圆对称纯位相片精密化设计,并发展了准梯度下降法及频谱滤波,改进了原有的混合杂化算法.模拟结果表明,考虑到热传导匀滑之后输出振幅分布顶部调制为1 1%,衍射效率为97 2%.Numerical Hankel transform for circularly symmetric system at variable sampling intervals in input and optput planes is presented. An improved hybrid algorithm by using quasi Gradient Decent algorithm and spectrum selectivity is applied to the design of pure phase elements (PPE) in uniform illumination. Considering heat conductivity the simulated result shows that 97\^2% of the incident beam energy is homogeneously concentrated into the desired focal spot in the target plane, and it fits to the required profile with the top profile error 1\^1%.

关 键 词：变间距采样 准梯度下降法 光束整形 纯位相片 频谱滤波 惯性约束核聚变 衍射效率 

分 类 号：TL632[核科学技术—核技术及应用] O433[机械工程—光学工程]