基于混合微粒群算法的相位解缠

PHASE UNWRAPPING WITH HYBRID PARTICLE SWARM OPTIMIZATION ALGORITHM

针对Goldste in枝切线及蚁群解缠方法存在的问题,提出一种基于混合微粒群算法的相位解缠方法。该算法首先对奇异点进行预处理,将其分为偶极点、边界点和内部点;然后,利用相干图对内部点分区,并采用混合微粒群算法生成各自区域内的最短路径;最后,将这些最短路径分割成电荷平衡的小段枝切线,并对边界点和剩余未连接的奇异点按照窗口扩展法生成枝切线。实验表明,该算法比传统枝切线解缠方法更加有效。

Aiming at the disadvantages of phase unwrapping methods for the Goldstein branch cutting and ant colony optimization,a new phase unwrapping algorithm with hybrid particle swarm optimization algorithm（H-PSO） is proposed.At first,the residues were divided into dipoles,border points and inner points by preprocessing,and then,the inner points were separated in several areas,where the shortest paths were generated by the H-PSO,finally,the shortest paths were segmented into some small branch cuts with the balance charges,and the border points and the disconnected residues were connected with window expanding method.The experimental results show that the algorithm proposed is much better than the classical phase unwrapping method of Goldstein branch cut.