摘要
针对位相差异(Phase Diversity,PD)技术存在计算量大和神经网络训练样本集规模过大等缺点,提出一种结合人工神经网络的PD波前传感新方法-基于自组织映射(Self-Organizing Featwe Map,SOFM)网络的PD波前传感方法.通过计算机仿真实验对新方法的性能进行了测试和分析.仿真实验结果表明,该方法在保证传感准确度的前提下可以有效提高PD波前传感的速度和缩减网络训练样本集的规模.
A method based on the Combination of Phase Diversity(CPD) with neural networks(NN) was proposed,which is Phase Diversity Wave-front Sensing based on Seer-organizing Feature Map (SOFM) NN. The characteristic of SOFM NN and the basic principle of this method were introduced. The performance of this method was evaluated and analysed by computer simulation. The simulation results show that this method can improve the speed of PD Wave-front Sensing and reduce the size of training set effectively. It has a potential value of practical application.
出处
《光子学报》
EI
CAS
CSCD
北大核心
2008年第7期1373-1377,共5页
Acta Photonica Sinica
关键词
稀疏孔径
波前传感
神经网络
Spare aperture
Wave-front sensing
Neural network